From nginx-forum at nginx.us  Sun May  1 05:04:21 2011
From: nginx-forum at nginx.us (inspire22)
Date: Sat, 30 Apr 2011 21:04:21 -0400
Subject: variables in if regexes
Message-ID: <96751c7eeff6de7ab97f1c1904c619bb.NginxMailingListEnglish@forum.nginx.org>

I have a caching system I need to bypass if the logging in user's name
is found in the $request_uri.  I'm trying to do something like this, but
can't get the variable to interpolate into the regex.  Any suggestions
pretty please?

    set $me "Asabouros.";    
    if ($uri ~ $me) { set $chk ""; }    

I've always tried things like this:
    if ($uri ~ "by-{$me}") { set $chk ""; }    

Thanks!
-Kevin

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,194753,194753#msg-194753



From wendal1985 at gmail.com  Sun May  1 05:28:16 2011
From: wendal1985 at gmail.com (Wendal Chen)
Date: Sun, 1 May 2011 09:28:16 +0800
Subject: nginx and google
In-Reply-To: <4daa0cf8766acfaaccec44582e1d2037.NginxMailingListEnglish@forum.nginx.org>
References: <4daa0cf8766acfaaccec44582e1d2037.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTi=MW6K-YAcqadZi64G0Lg1xjsscDA@mail.gmail.com>

check your robot.txt

2011/4/30 konelli <nginx-forum at nginx.us>

> freebsd 8 nginx-0.8.54 proxy works perfect. Any site loads very fast
> thru proxy pass  but google do not wont to load! Any other indexing
> sites. Help.
> Here my config
>
> worker_processes  1;
> events {
>    worker_connections  5;
> }
>
> http {
>          server {
>              listen 8081;
> #             access_log off;
>
>        location / {
>           proxy_pass http://211.141.153.146;
>           proxy_set_header Host $host;
>        }
>     }
> }
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,194580,194580#msg-194580
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>



-- 
Wendal Chen
<http://t.sina.com.cn/1068040557?s=6uyXnP>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110501/c76ee77c/attachment.html>

From nginx-forum at nginx.us  Mon May  2 04:26:31 2011
From: nginx-forum at nginx.us (Andy)
Date: Sun, 01 May 2011 20:26:31 -0400
Subject: Does Nginx block on file IO?
Message-ID: <f8e0975a0299f5264a2f49ddb6ac1153.NginxMailingListEnglish@forum.nginx.org>

Hi.

I was always under the impression that Nginx is non-blocking for file
IO. Then I was told it wasn't.

I'm considering using Nginx to serve static images. Pretty much every
connection will result in a file IO. If Nginx blocks for file IO, then
using Nginx here wouldn't be any better than using Apache, right? Every
connection will lead to a file IO which blocks the entire Nginx process.
So to serve 500 concurrent connections I'll need 500 Nginx processes.

Would Nginx work in such a use case? Any tips on how to use Nginx for
serving static media?

Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,194884,194884#msg-194884



From sweeks at imvu.com  Mon May  2 04:54:17 2011
From: sweeks at imvu.com (Stephen Weeks)
Date: Sun, 1 May 2011 17:54:17 -0700
Subject: Crash (double free or corruption) on trying to proxy to localhost
Message-ID: <BANLkTikAQ-FGHBUT4+eFiEQ2stY88f10ow@mail.gmail.com>

I've currently got a pool of systems running nginx proxying to a pool of
systems running apache.  I'm trying to move to running nginx locally on the
apache hosts instead of separate systems to avoid the extra network hop,
make more-efficient use of resources, and enable some future development
(including migrating to running our application on nginx via fastcgi instead
of apache, ideally).  we've currently got some significant architecture
built up around apache, so converting right now is uncomfortable.

Ideally, I'd like nginx to just serve from localhost, but fail over to the
rest of the pool when localhost in unavailable, so in my upstream I have
every server except for localhost set as 'backup'.  I'm otherwise running
identical configurations of apache and nginx on a single system together as
used in the rest of the two pools.  This works exactly as expected, except
that I get a few crashes of nginx workers every minute.  This only happens
when proxying to the local system.  If I proxy anywhere else, it works
fine.  Other proxies can serve from this system without trouble.  I see this
same behaviour on other hosts when I build them the same way, so it's not an
error with the host.  I see this crash on 0.7.65, 0.8.54, and 1.0.0, running
on Ubuntu 10.04 LTS.  I see this crash whether I'm connecting to 127.0.0.1
or the host's local IP.  I see this crash whether I'm listening on *:80 or
<public ip>:80.  I see this crash whether I'm connecting to :80 or running
apache on a different port and connecting to :81.  I see this crash whether
I'm running ubuntu's "nginx-light" configuration, or their "nginx-full"
configuration.  I see no errors logged from apache.

1) I'd really love to make this work, so if there's anything else I can try,
any additional debugging information I can give, I'd appreciate it.
2) Nginx has been very useful to me so far, so I thought you'd appreciate a
bug report.

Posted on github, I have a problem description, section of a debug log, my
(slightly edited: flattened includes and stripped an IP) nginx.conf, a gdb
backtrace, and some additional information I was asked for when looking for
help on IRC.  This is everything I've been able to come up with that sounds
plausibly relevant.

https://gist.github.com/1574dbaf3a3dcda920a2

Any help?

(Thanks to MTechnology and kolbyjack for helping me troubleshoot this on
IRC)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110501/9d18eb10/attachment.html>

From nginx-forum at nginx.us  Mon May  2 05:54:27 2011
From: nginx-forum at nginx.us (jjjx128)
Date: Sun, 01 May 2011 21:54:27 -0400
Subject: Does Nginx block on file IO?
In-Reply-To: <f8e0975a0299f5264a2f49ddb6ac1153.NginxMailingListEnglish@forum.nginx.org>
References: <f8e0975a0299f5264a2f49ddb6ac1153.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <b32cbe4d8f40f62d5dbed302cf0ff3b0.NginxMailingListEnglish@forum.nginx.org>

Andy Wrote:
-------------------------------------------------------
> I was always under the impression that Nginx is
> non-blocking for file IO. Then I was told it
> wasn't.

Actually AIO support was added not so long ago. It can be non-blocking
for file IO on systems with AIO. 

> If Nginx blocks for file IO, then
> using Nginx here wouldn't be any better than using
> Apache, right? 

How is it even related? There is a difference between getting a file
from disk and serving it to the client from buffer. You can get it
really fast and block for a very little time but than send it to the
client very slowly. So asynchronous front-end would be a really good
idea. 
Other than that you might find that neither nginx nor apache can serve
your images fast enough from any common filesystem. You will need
something like Bitcask to go farther. The idea is to get an entire image
to the buffer in a single disk seek.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,194884,194891#msg-194891



From nginx-forum at nginx.us  Mon May  2 06:59:51 2011
From: nginx-forum at nginx.us (flubber)
Date: Sun, 01 May 2011 22:59:51 -0400
Subject: nginx and php fast cgi different server?
Message-ID: <f2f374c76ba5d1723d5ac9b149553924.NginxMailingListEnglish@forum.nginx.org>

Hello All,

I'm a newbie on nginx, currently i have configure nginx n php fast cgi
with same server, i want to split nginx and php fast cgi with different,
it's posibble?

I have test change fastcgi_pass point di php fast cgi to other server,
but when access site just get "No input file specified." and no error
report.

This my configuration.

        location ~ \.php$ {
                #fastcgi_pass   127.0.0.1:9000;
                fastcgi_pass   10.1.1.10:9000;
                fastcgi_index  index.php;
                fastcgi_param  SCRIPT_FILENAME 
/var/www/html$fastcgi_script_name;
                include fastcgi_params;
        }

My nginx nginx/0.9.6, please advice and suggestion

Thx all.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,194899,194899#msg-194899



From edhoprima at gmail.com  Mon May  2 07:02:49 2011
From: edhoprima at gmail.com (Edho P Arief)
Date: Mon, 2 May 2011 10:02:49 +0700
Subject: nginx and php fast cgi different server?
In-Reply-To: <f2f374c76ba5d1723d5ac9b149553924.NginxMailingListEnglish@forum.nginx.org>
References: <f2f374c76ba5d1723d5ac9b149553924.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTi=aTmg1dTgAyO5DyMgGfW3Po-gn+w@mail.gmail.com>

On Mon, May 2, 2011 at 9:59 AM, flubber <nginx-forum at nginx.us> wrote:
> Hello All,
>
> I'm a newbie on nginx, currently i have configure nginx n php fast cgi
> with same server, i want to split nginx and php fast cgi with different,
> it's posibble?
>
> I have test change fastcgi_pass point di php fast cgi to other server,
> but when access site just get "No input file specified." and no error
> report.
>
> This my configuration.
>
> ? ? ? ?location ~ \.php$ {
> ? ? ? ? ? ? ? ?#fastcgi_pass ? 127.0.0.1:9000;
> ? ? ? ? ? ? ? ?fastcgi_pass ? 10.1.1.10:9000;
> ? ? ? ? ? ? ? ?fastcgi_index ?index.php;
> ? ? ? ? ? ? ? ?fastcgi_param ?SCRIPT_FILENAME
> /var/www/html$fastcgi_script_name;
> ? ? ? ? ? ? ? ?include fastcgi_params;
> ? ? ? ?}
>
> My nginx nginx/0.9.6, please advice and suggestion
>

the fastcgi files (in this case, php files) must be located in the
other server (10.1.1.10 in this case)


From sweeks at imvu.com  Mon May  2 07:13:42 2011
From: sweeks at imvu.com (Stephen Weeks)
Date: Sun, 1 May 2011 20:13:42 -0700
Subject: Crash (double free or corruption) on trying to proxy to localhost
In-Reply-To: <BANLkTikAQ-FGHBUT4+eFiEQ2stY88f10ow@mail.gmail.com>
References: <BANLkTikAQ-FGHBUT4+eFiEQ2stY88f10ow@mail.gmail.com>
Message-ID: <BANLkTimP_foWzJeM+a8hNZL+QhF-FnBxAw@mail.gmail.com>

Apparently these crashed don't happen with the upstreamfair module, as far
as I've been able to tell so far.  That's probably good enough as a
workaround for me for now, and may be a clue when investigating this crash.

On Sun, May 1, 2011 at 5:54 PM, Stephen Weeks <sweeks at imvu.com> wrote:

> I've currently got a pool of systems running nginx proxying to a pool of
> systems running apache.  I'm trying to move to running nginx locally on the
> apache hosts instead of separate systems to avoid the extra network hop,
> make more-efficient use of resources, and enable some future development
> (including migrating to running our application on nginx via fastcgi instead
> of apache, ideally).  we've currently got some significant architecture
> built up around apache, so converting right now is uncomfortable.
>
> Ideally, I'd like nginx to just serve from localhost, but fail over to the
> rest of the pool when localhost in unavailable, so in my upstream I have
> every server except for localhost set as 'backup'.  I'm otherwise running
> identical configurations of apache and nginx on a single system together as
> used in the rest of the two pools.  This works exactly as expected, except
> that I get a few crashes of nginx workers every minute.  This only happens
> when proxying to the local system.  If I proxy anywhere else, it works
> fine.  Other proxies can serve from this system without trouble.  I see this
> same behaviour on other hosts when I build them the same way, so it's not an
> error with the host.  I see this crash on 0.7.65, 0.8.54, and 1.0.0, running
> on Ubuntu 10.04 LTS.  I see this crash whether I'm connecting to 127.0.0.1
> or the host's local IP.  I see this crash whether I'm listening on *:80 or
> <public ip>:80.  I see this crash whether I'm connecting to :80 or running
> apache on a different port and connecting to :81.  I see this crash whether
> I'm running ubuntu's "nginx-light" configuration, or their "nginx-full"
> configuration.  I see no errors logged from apache.
>
> 1) I'd really love to make this work, so if there's anything else I can
> try, any additional debugging information I can give, I'd appreciate it.
> 2) Nginx has been very useful to me so far, so I thought you'd appreciate a
> bug report.
>
> Posted on github, I have a problem description, section of a debug log, my
> (slightly edited: flattened includes and stripped an IP) nginx.conf, a gdb
> backtrace, and some additional information I was asked for when looking for
> help on IRC.  This is everything I've been able to come up with that sounds
> plausibly relevant.
>
> https://gist.github.com/1574dbaf3a3dcda920a2
>
> Any help?
>
> (Thanks to MTechnology and kolbyjack for helping me troubleshoot this on
> IRC)
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110501/5ba77771/attachment.html>

From lists at ruby-forum.com  Mon May  2 08:08:24 2011
From: lists at ruby-forum.com (Mike P.)
Date: Mon, 02 May 2011 06:08:24 +0200
Subject: 502 errors, upstream sent too big header
In-Reply-To: <BANLkTi=CkxwWv5aW8F+Rxn6DCanaebLnhQ@mail.gmail.com>
References: <BANLkTi=CkxwWv5aW8F+Rxn6DCanaebLnhQ@mail.gmail.com>
Message-ID: <79524ba8fb1d2c8fb975fab30434c302@ruby-forum.com>

Hi,

I'm getting this same error (please see log output below) with
Nginx(0.8.54) and Passenger(3.0.6) using GET requests with really long
URIs.

My assumption is that fastcgi_buffer* directives won't apply to my
configuration since the request is being handled by passenger instead of
a php/perl fastcgi socket.

I've tried defining the proxy_buffer* settings like Travis described in
his initial post, but that didn't resolve the error I'm experiencing.

I've defined "large_client_header_buffers 8 16k;"  in the server
context, but this doesn't alleviate the error either.

Is 'large_client_header_buffers' the proper directive I should be using?
If not, is there a directive I can use to increase the response header
size limit?

Thanks for any insight into this,
Mike P.

----------
2011/05/01 09:57:41 [error] 12176#0: *4537092 upstream sent too big
header while reading response header from upstream, client: x.x.x.x,
server: hostname.tld, request: "GET
/heroku/resources/18623?<REALLY_LONG_URI_REMOVED> HTTP/1.1", upstream:
"passenger:unix:/
----------

-- 
Posted via http://www.ruby-forum.com/.


From nginx-forum at nginx.us  Mon May  2 08:11:10 2011
From: nginx-forum at nginx.us (HowToMeetLadies)
Date: Mon, 02 May 2011 00:11:10 -0400
Subject: Catch ALL requests by LUA-script
In-Reply-To: <eb9c211f9f74635a1e0afcc9d0591e22.NginxMailingListEnglish@forum.nginx.org>
References: <AANLkTimcqUR5y2Yi2e4B5j5k+yZ6GDOvxRmT-Qm4M2NH@mail.gmail.com>
	<eb9c211f9f74635a1e0afcc9d0591e22.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <90c728b1bd4f045a01171963ba75b30f.NginxMailingListEnglish@forum.nginx.org>

Maybe slightly off-topic but then maybe it is useful for you (andiL) or
someone else.

Before i start, @agentzh: every ?nth request, one of nginx
(0.8.54.4rc2-jit2beta6) worker processes segfaults, is this behaviour
related to the pcre issue? have not patched it yet.. ah and thank you
*very* much for giving nginx this all new powers!! ((:

Ok, im a long time nginx/redis/ruby user and this are my first steps
with lua and a lightweight (M)VC framework around three core "classes",
which uses haml for templates/layout and markdown for text processing (5
minute ffi discount binding), drizzle and redis for data and redis also
for caching. Successfully running a little app with it. yay! *.* For now
i can give you some snippets just to show some basic aspects, ill
release a fully working example sometime soon, if the segfaults doesnt
belong to the pcre-issue ill try to debug that too.


Below ill use luasql, instead you can use
ngx.location.capture("/some-internal-rds-json") and lua-yajl, but this
approach ends as soon as you want INSERT followed by LAST_INSERT_ID() or
other complex queries/transactions (as long as libdrizzle has no support
for multiple queries, maybe this is not the best option to go with, ive
tried alternatives and workarounds, but for now this is what *I* want).

At the moment i do not need an ORM for my data, but have already done
some functions for both rds_json and luasql. redis has some commands
exposed in nginx.conf for easy access with lua (: The core classes:

Request; which contains several ngx.var wrappers (for nil instead of ''
for empty keys), present(str), simple cookie/session/login, before/after
filters, haml/markdown rendering, etc..

MethodRequest < Request; which calls functions for HTTP Methods aka GET,
POST alike..
and finally Resource < MethodRequest; which has an
Rails-ActionController-like interface.


Greetz
# HTML

--- ATTACHMENT ---

ookeh to bad no code formatting...?

http://exc.eu/paste/deadbeaf.txt

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,179364,194903#msg-194903



From nginx-forum at nginx.us  Mon May  2 08:26:29 2011
From: nginx-forum at nginx.us (flubber)
Date: Mon, 02 May 2011 00:26:29 -0400
Subject: nginx and php fast cgi different server?
In-Reply-To: <BANLkTi=aTmg1dTgAyO5DyMgGfW3Po-gn+w@mail.gmail.com>
References: <BANLkTi=aTmg1dTgAyO5DyMgGfW3Po-gn+w@mail.gmail.com>
Message-ID: <037843eb881f9c7979186de05bb8d829.NginxMailingListEnglish@forum.nginx.org>

it's works, thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,194899,194905#msg-194905



From nginx-forum at nginx.us  Mon May  2 09:15:08 2011
From: nginx-forum at nginx.us (moorthi)
Date: Mon, 02 May 2011 01:15:08 -0400
Subject: nginx imaps auth_http cyrus
In-Reply-To: <3f29c76ca3fec62ea3f4f3423b70de4e.NginxMailingListEnglish@forum.nginx.org>
References: <4D73FB1E.3090104@alokat.org>
	<3f29c76ca3fec62ea3f4f3423b70de4e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <8599d7049c20a5f08721167b39775a44.NginxMailingListEnglish@forum.nginx.org>

auth_http script and other setttings are done correctly,
this is what i am getting in error.log
upstream timed out (111: Connection timed out) while connecting to
upstream
it's saying +OK for user <username> in connect
but after entering pass <password> it's giving above error.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,180902,194908#msg-194908



From piotr.sikora at frickle.com  Mon May  2 09:22:52 2011
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Mon, 2 May 2011 07:22:52 +0200
Subject: Does Nginx block on file IO?
In-Reply-To: <f8e0975a0299f5264a2f49ddb6ac1153.NginxMailingListEnglish@forum.nginx.org>
References: <f8e0975a0299f5264a2f49ddb6ac1153.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <A5A1B25BFEEC4DD388CE6118B59F28E6@Desktop>

Hi,

> I was always under the impression that Nginx is non-blocking for file
> IO. Then I was told it wasn't.

That depends on your configuration on on capabilities of your operating 
system. Asynchronous file I/O is available on recent versions of Linux, 
Solaris, DragonFly, FreeBSD, NetBSD, MacOSX.

> I'm considering using Nginx to serve static images. Pretty much every
> connection will result in a file IO.

No, it won't. Recently served files should be in your OS's buffer cache.

> If Nginx blocks for file IO, then
> using Nginx here wouldn't be any better than using Apache, right? Every
> connection will lead to a file IO which blocks the entire Nginx process.
> So to serve 500 concurrent connections I'll need 500 Nginx processes.

If that would be the case (buffer cache hit rate would be very low) and you 
would be using magnetic disks, then yes, you'd be better off with a web 
server than uses per-client lightweight processes or threads.

Best regards,
Piotr Sikora < piotr.sikora at frickle.com >



From piotr.sikora at frickle.com  Mon May  2 09:29:30 2011
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Mon, 2 May 2011 07:29:30 +0200
Subject: Does Nginx block on file IO?
In-Reply-To: <b32cbe4d8f40f62d5dbed302cf0ff3b0.NginxMailingListEnglish@forum.nginx.org>
References: <f8e0975a0299f5264a2f49ddb6ac1153.NginxMailingListEnglish@forum.nginx.org>
	<b32cbe4d8f40f62d5dbed302cf0ff3b0.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <C5FEDA6D51354B32A67742AFF4764E4B@Desktop>

Hi,

> How is it even related? There is a difference between getting a file
> from disk and serving it to the client from buffer. You can get it
> really fast and block for a very little time but than send it to the
> client very slowly.

It's actually very related, because with event-driven programming, you're 
blocking _all_ worker's clients on the single disk I/O.

Best regards,
Piotr Sikora < piotr.sikora at frickle.com >



From justin at specialbusservice.com  Mon May  2 15:06:41 2011
From: justin at specialbusservice.com (Justin Cormack)
Date: Mon, 02 May 2011 12:06:41 +0100
Subject: Does Nginx block on file IO?
In-Reply-To: <C5FEDA6D51354B32A67742AFF4764E4B@Desktop>
References: <f8e0975a0299f5264a2f49ddb6ac1153.NginxMailingListEnglish@forum.nginx.org>
	<b32cbe4d8f40f62d5dbed302cf0ff3b0.NginxMailingListEnglish@forum.nginx.org>
	<C5FEDA6D51354B32A67742AFF4764E4B@Desktop>
Message-ID: <1304334401.2438.145.camel@mackerel>

On Mon, 2011-05-02 at 07:29 +0200, Piotr Sikora wrote:
> Hi,
> 
> > How is it even related? There is a difference between getting a file
> > from disk and serving it to the client from buffer. You can get it
> > really fast and block for a very little time but than send it to the
> > client very slowly.
> 
> It's actually very related, because with event-driven programming, you're 
> blocking _all_ worker's clients on the single disk I/O.

If you expect file IO to block, use more nginx processes than cores. I
dont think that async IO on Linux is really worth using, although I will
do some benchmarks. You might try switching to FreeBSD as well. 

I think that node.js currently uses a thread pool for async IO on Linux
(possibly all platforms, except maybe Windows), with sync IO in the
background. Linux kernel aio was designed for programs like databases
that manage their own buffer caches, not for normal file IO. If you dont
cache the results it will be slow (unless anything has changed recently,
will do some testing).

Justin




From nginx-forum at nginx.us  Mon May  2 15:17:08 2011
From: nginx-forum at nginx.us (jjjx128)
Date: Mon, 02 May 2011 07:17:08 -0400
Subject: Does Nginx block on file IO?
In-Reply-To: <C5FEDA6D51354B32A67742AFF4764E4B@Desktop>
References: <C5FEDA6D51354B32A67742AFF4764E4B@Desktop>
Message-ID: <a1990723ac5b22baa79483a8bd519294.NginxMailingListEnglish@forum.nginx.org>

Piotr Sikora Wrote:
-------------------------------------------------------
> It's actually very related, because with
> event-driven programming, you're 
> blocking _all_ worker's clients on the single disk
> I/O.

I know, but imagine how much memory you can save and use for file system
cache if you are not running apache. Which can make fat more greater
impact on your performance. You can't just say that nginx is the same as
apache just because it's blocking on file I/O. 
So still, it depends on the type of load you are having. And if you are
not serving 100 megabytes of data to 500 clients but a lot more you are
most likely to get better results with asynchronous server even with
blocking file I/O.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,194884,194939#msg-194939



From nginx-forum at nginx.us  Mon May  2 15:25:54 2011
From: nginx-forum at nginx.us (jjjx128)
Date: Mon, 02 May 2011 07:25:54 -0400
Subject: Does Nginx block on file IO?
In-Reply-To: <a1990723ac5b22baa79483a8bd519294.NginxMailingListEnglish@forum.nginx.org>
References: <C5FEDA6D51354B32A67742AFF4764E4B@Desktop>
	<a1990723ac5b22baa79483a8bd519294.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <8343634767e2d40d5164389d05e4884f.NginxMailingListEnglish@forum.nginx.org>

jjjx128 Wrote:
-------------------------------------------------------
> And if you are not serving 100 megabytes
> of data to 500 clients but a lot more you are most
> likely to get better results with asynchronous
> server even with blocking file I/O.

Sorry, bad example, in this case your data will fit in file system cache
and you still will be better off with nginx bacuase of all the context
switches you save. Anyway there might be some very special case where
disks are overloaded and access patterns are truly random. Although this
never happens in real world.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,194884,194941#msg-194941



From markus.jelsma at openindex.io  Mon May  2 16:56:19 2011
From: markus.jelsma at openindex.io (Markus Jelsma)
Date: Mon, 2 May 2011 14:56:19 +0200
Subject: Using map and proxy_pass
In-Reply-To: <201104301252.43466.markus.jelsma@openindex.io>
References: <201104291748.48357.markus.jelsma@openindex.io>
	<f975fd982b56ba3953d51b7e1892d895.NginxMailingListEnglish@forum.nginx.org>
	<201104301252.43466.markus.jelsma@openindex.io>
Message-ID: <201105021456.19711.markus.jelsma@openindex.io>

Anyone else having some thoughs on how to use the map in conjunction with 
proxy_pass to set different hosts based on the first URI segment?

On Saturday 30 April 2011 12:52:42 Markus Jelsma wrote:
> Too many. I need to generate the mappings in an external program. I'd
> rather not generate as many different locations.
> 
> 
> Thanks
> 
> > How many uri / backend mappings do you have?  You could perform the
> > above in two straight forward location blocks, based on the uri.
> > 
> > location /abc {
> > 
> > proxy_pass http://host01:8080;
> > }
> > 
> > ....
> > 
> > Posted at Nginx Forum:
> > http://forum.nginx.org/read.php?2,194480,194616#msg-194616
> > 
> > 
> > _______________________________________________
> > nginx mailing list
> > nginx at nginx.org
> > http://nginx.org/mailman/listinfo/nginx
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx

-- 
Markus Jelsma - CTO - Openindex
http://www.linkedin.com/in/markus17
050-8536620 / 06-50258350


From nginx-forum at nginx.us  Mon May  2 18:22:50 2011
From: nginx-forum at nginx.us (monster)
Date: Mon, 02 May 2011 10:22:50 -0400
Subject: How to get a 404 logged with try_files?
Message-ID: <eece71dda665fc6cf671ec665fe4db89.NginxMailingListEnglish@forum.nginx.org>

Hi! Just started with Nginx after reading it reached 1.0.
Congratulation!

I'm trying to get my test site to return a "missing image" icon when an
image is missing, but have a 404 logged so I know I messed up something.
I want to specify the images name without file extensions, in case I
move some from PNG to JPG, or the other way around. My problem is that
try_files doc says last path MUST exist. I would like last path to be
$uri and cause a 404, not a 500. Tried using error_page to convert 500
to 404, but did not have any effect. Thanks for any advice!

location ~ ^/img/[^\.]*$ {
    try_files $uri.png $uri.jpg $uri.gif $uri;
    error_page  500 =404	/img/missing;
}

location /img/ {
    error_page  404		/img/missing;
}

location = /img/missing {
    rewrite .* /img/icon/missing.png;
}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,194965,194965#msg-194965



From francis at daoine.org  Mon May  2 19:35:31 2011
From: francis at daoine.org (Francis Daly)
Date: Mon, 2 May 2011 16:35:31 +0100
Subject: Using map and proxy_pass
In-Reply-To: <201105021456.19711.markus.jelsma@openindex.io>
References: <201104291748.48357.markus.jelsma@openindex.io>
	<f975fd982b56ba3953d51b7e1892d895.NginxMailingListEnglish@forum.nginx.org>
	<201104301252.43466.markus.jelsma@openindex.io>
	<201105021456.19711.markus.jelsma@openindex.io>
Message-ID: <20110502153531.GB12564@craic.sysops.org>

On Mon, May 02, 2011 at 02:56:19PM +0200, Markus Jelsma wrote:

Hi there,

A general rule for most config confusions is "the debug log is your
friend".

> Anyone else having some thoughs on how to use the map in conjunction with 
> proxy_pass to set different hosts based on the first URI segment?

It's not clear to me why you're happy to generate an include file of
mappings in an external program, but not happy to generate an include
file of locations in an external program.

And if you have a large number of locations, you'll probably be happier
having no top-level regex locations.

But all that aside...

"map $var1 $var2" sets $var2 based on a string match of $var1.

You wish to match based on the first uri segment. So, do just that:

===
http {
  map $first_segment $value {
      default     127.0.0.1;
      /abc        127.0.0.2;
      /def        127.0.0.3;
  }
  server {
    if ($uri ~* ^(/[^/]*) ) {
      set $first_segment $1;
    }
    location / {
      proxy_pass http://$value:8080/$uri ;
    }
  }
}
===

You may want to play with the proxy_pass line depending on exactly what
you want to do, but the above seems to work for me.

You can also add a $second_segment (or $part2) in the if{} block, if
you want a separate map for that.

Good luck with it,

	f
-- 
Francis Daly        francis at daoine.org


From lists at ruby-forum.com  Mon May  2 21:54:57 2011
From: lists at ruby-forum.com (Yanxin Z.)
Date: Mon, 02 May 2011 19:54:57 +0200
Subject: nginx rewrite
Message-ID: <721e551f14c3665cad91466a13a63001@ruby-forum.com>

Hello,
I need to rewrite the URL

/img/14567.jpg to /img/14/14567.jpg

The image name is 14567.jpg, which is saved in /14 folder. /14 is
caluculated by 14567 / 1000

How to implement is NGX rewrite?

Thanks,
Yanxin

-- 
Posted via http://www.ruby-forum.com/.


From cliff at develix.com  Mon May  2 22:11:04 2011
From: cliff at develix.com (Cliff Wells)
Date: Mon, 02 May 2011 11:11:04 -0700
Subject: Does Nginx block on file IO?
In-Reply-To: <f8e0975a0299f5264a2f49ddb6ac1153.NginxMailingListEnglish@forum.nginx.org>
References: <f8e0975a0299f5264a2f49ddb6ac1153.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <1304359864.11635.47.camel@portable-evil>

On Sun, 2011-05-01 at 20:26 -0400, Andy wrote:
> Hi.
> 
> I was always under the impression that Nginx is non-blocking for file
> IO. Then I was told it wasn't.
> 
> I'm considering using Nginx to serve static images. Pretty much every
> connection will result in a file IO. If Nginx blocks for file IO, then
> using Nginx here wouldn't be any better than using Apache, right? Every
> connection will lead to a file IO which blocks the entire Nginx process.
> So to serve 500 concurrent connections I'll need 500 Nginx processes.

You cannot take a single measure such as concurrent requests independent
of requests per second and make a good prediction about performance
requirements.  If you have 500 concurrent requests, and each request
takes an average of 1ms, then you could conceivably serve all 500
requests within 0.5 seconds with a single worker, which is quite
reasonable.  People have reported serving 10K requests per second using
Nginx on rather modest hardware (laptops even).  You'll probably find
that your OS is the bottleneck, not Nginx (and you should focus your
performance tuning at that level - TCP buffers and the like).

It's also worth pointing out that if you currently have 500 concurrent
requests on Apache, that number may actually decrease with Nginx,
assuming requests are finished faster (i.e. if Nginx finishes each
request twice as fast as Apache, you'd only see 250 concurrent requests,
while requests per second would double).  Obviously that's simplified,
but hopefully clarifies my point.

In any case, no need to have a worker per request, just start with the
recommendation of 1 worker per core and tune from there.  It's true that
the *concurrent* number of requests is limited by the number of workers
you have (the remaining requests will be served serially as each worker
is freed), but in practice, it becomes academic. Serialization is going
to happen at some level, even if you have a threaded server, so
basically software threads buy you little more than the feeling of
concurrency at the cost of massive amounts of memory.

> Would Nginx work in such a use case? Any tips on how to use Nginx for
> serving static media?

Serving static content is actually where Nginx has been demonstrated to
outperform Apache by a wide margin, in no small part because it leaves
lots of memory for filesystem caching, and also because it causes less
context-switches than a threaded server.

At the end of the day, you should do some performance testing, since
performance is going to depend a lot on factors that will only be
revealed on your particular setup and your particular data (number and
size of files, cache misses, etc).

Regards,
Cliff




From igor at sysoev.ru  Mon May  2 22:11:31 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Mon, 2 May 2011 22:11:31 +0400
Subject: nginx rewrite
In-Reply-To: <721e551f14c3665cad91466a13a63001@ruby-forum.com>
References: <721e551f14c3665cad91466a13a63001@ruby-forum.com>
Message-ID: <20110502181131.GF43927@sysoev.ru>

On Mon, May 02, 2011 at 07:54:57PM +0200, Yanxin Z. wrote:
> Hello,
> I need to rewrite the URL
> 
> /img/14567.jpg to /img/14/14567.jpg
> 
> The image name is 14567.jpg, which is saved in /14 folder. /14 is
> caluculated by 14567 / 1000
> 
> How to implement is NGX rewrite?

location /img/ {
    location ~ ^/img/(..)(.+)$ {
        alias   /path/to/img/$1/$1$2;
    }
}


-- 
Igor Sysoev


From nginx-forum at nginx.us  Mon May  2 22:21:48 2011
From: nginx-forum at nginx.us (monster)
Date: Mon, 02 May 2011 14:21:48 -0400
Subject: How to get a 404 logged with try_files?
In-Reply-To: <eece71dda665fc6cf671ec665fe4db89.NginxMailingListEnglish@forum.nginx.org>
References: <eece71dda665fc6cf671ec665fe4db89.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <e260c038e3d0be6bf8531e7e12252410.NginxMailingListEnglish@forum.nginx.org>

Found it! It must look like this, putting =404 as last location in
try_files:

location ~ ^/img/[^\.]*$ {
        try_files $uri.png $uri.jpg $uri.gif $uri =404;
        error_page  404	/img/missing;
}

That would be useful information to put in the try_files wiki...

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,194965,195019#msg-195019



From lists at ruby-forum.com  Mon May  2 22:48:50 2011
From: lists at ruby-forum.com (Yanxin Z.)
Date: Mon, 02 May 2011 20:48:50 +0200
Subject: nginx rewrite
In-Reply-To: <721e551f14c3665cad91466a13a63001@ruby-forum.com>
References: <721e551f14c3665cad91466a13a63001@ruby-forum.com>
Message-ID: <c37c4c0db9d72cf269f9a1d9dd3bb2ab@ruby-forum.com>

Hello Igor Sysoev,
Thank you for your help.

You suggestion is great, but my problem is a little bit complicated.

The image name is flexible, can be from 0.jpg to 23456.jpg, we organize 
the image files into folders, whose name is the number divide by 1000. 
For example, the folder 100 saves the file 100000.jpg to 100999.jpg.

Both the folder and file name are flexible in my case.

Do you have any suggestion on it?

Thanks,
Yanxin

-- 
Posted via http://www.ruby-forum.com/.


From mdounin at mdounin.ru  Mon May  2 22:53:27 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 2 May 2011 22:53:27 +0400
Subject: nginx imaps auth_http cyrus
In-Reply-To: <8599d7049c20a5f08721167b39775a44.NginxMailingListEnglish@forum.nginx.org>
References: <4D73FB1E.3090104@alokat.org>
	<3f29c76ca3fec62ea3f4f3423b70de4e.NginxMailingListEnglish@forum.nginx.org>
	<8599d7049c20a5f08721167b39775a44.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110502185327.GD42265@mdounin.ru>

Hello!

On Mon, May 02, 2011 at 01:15:08AM -0400, moorthi wrote:

> auth_http script and other setttings are done correctly,
> this is what i am getting in error.log
> upstream timed out (111: Connection timed out) while connecting to
> upstream

I believe this is clear enough: nginx wasn't able to connect to 
upstream (backend) server.  Check backend server and port returned 
by auth_http script.

> it's saying +OK for user <username> in connect
> but after entering pass <password> it's giving above error.

After PASS command nginx will try to authenticate client with 
auth_http script and then connect to upstream (backend) server.  
As soon as connection to upstream times out - nginx should return 
"-ERR internal server error" in reply to PASS command (if client 
is still here).

Maxim Dounin


From mdounin at mdounin.ru  Mon May  2 22:57:03 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 2 May 2011 22:57:03 +0400
Subject: 502 errors, upstream sent too big header
In-Reply-To: <79524ba8fb1d2c8fb975fab30434c302@ruby-forum.com>
References: <BANLkTi=CkxwWv5aW8F+Rxn6DCanaebLnhQ@mail.gmail.com>
	<79524ba8fb1d2c8fb975fab30434c302@ruby-forum.com>
Message-ID: <20110502185703.GE42265@mdounin.ru>

Hello!

On Mon, May 02, 2011 at 06:08:24AM +0200, Mike P. wrote:

> Hi,
> 
> I'm getting this same error (please see log output below) with
> Nginx(0.8.54) and Passenger(3.0.6) using GET requests with really long
> URIs.
> 
> My assumption is that fastcgi_buffer* directives won't apply to my
> configuration since the request is being handled by passenger instead of
> a php/perl fastcgi socket.

You have to tune appropriate settings provided by passenger 
module.  Please refer to passenger documentation for exact names, 
most likely passenger_buffer_size and so on.

[...]

Maxim Dounin


From igor at sysoev.ru  Mon May  2 22:57:16 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Mon, 2 May 2011 22:57:16 +0400
Subject: nginx rewrite
In-Reply-To: <c37c4c0db9d72cf269f9a1d9dd3bb2ab@ruby-forum.com>
References: <721e551f14c3665cad91466a13a63001@ruby-forum.com>
	<c37c4c0db9d72cf269f9a1d9dd3bb2ab@ruby-forum.com>
Message-ID: <20110502185716.GG43927@sysoev.ru>

On Mon, May 02, 2011 at 08:48:50PM +0200, Yanxin Z. wrote:
> Hello Igor Sysoev,
> Thank you for your help.
> 
> You suggestion is great, but my problem is a little bit complicated.
> 
> The image name is flexible, can be from 0.jpg to 23456.jpg, we organize 
> the image files into folders, whose name is the number divide by 1000. 
> For example, the folder 100 saves the file 100000.jpg to 100999.jpg.
> 
> Both the folder and file name are flexible in my case.
> 
> Do you have any suggestion on it?

location /img/ {

    location ~ ^/img/(\d)(\.jpg)$ {
        alias   /path/to/img/0$1/$1$2;
    }

    location ~ ^/img/(\d\d)(\d*\.jpg)$ {
        alias   /path/to/img/$1/$1$2;
    }

    return 404;
}


-- 
Igor Sysoev


From nginx-forum at nginx.us  Mon May  2 23:04:48 2011
From: nginx-forum at nginx.us (dbanks)
Date: Mon, 02 May 2011 15:04:48 -0400
Subject: Shared nginx configuration for multiple servers
In-Reply-To: <BANLkTimxeg3-hW=9x0K7-Q+RdC2D5fBJAQ@mail.gmail.com>
References: <BANLkTimxeg3-hW=9x0K7-Q+RdC2D5fBJAQ@mail.gmail.com>
Message-ID: <de91070ac3595c06a46807fd3c68a3bb.NginxMailingListEnglish@forum.nginx.org>

I'm facing a related issue, although my attempt at implementation is
different from the OP.

ip.conf:
set $server_ip 10.11.1.50;

nginx.conf:
...
server { 
        include ip.conf;  #adds the $server_ip variable which is
machine-specific
        listen  $server_ip:80 default_server;
...
}

This fails with:  nginx: [emerg] host not found in "$server_ip:80" of
the "listen" directive in /opt/nginx/conf/nginx.conf:48

What is the best way to include a hard-coded IP address from another
file in order to allow nginx.conf to be universal across deployed
machines?

Cheers,

Dean

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,188621,195032#msg-195032



From mdounin at mdounin.ru  Mon May  2 23:19:06 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 2 May 2011 23:19:06 +0400
Subject: nginx rewrite
In-Reply-To: <20110502185716.GG43927@sysoev.ru>
References: <721e551f14c3665cad91466a13a63001@ruby-forum.com>
	<c37c4c0db9d72cf269f9a1d9dd3bb2ab@ruby-forum.com>
	<20110502185716.GG43927@sysoev.ru>
Message-ID: <20110502191906.GF42265@mdounin.ru>

Hello!

On Mon, May 02, 2011 at 10:57:16PM +0400, Igor Sysoev wrote:

> On Mon, May 02, 2011 at 08:48:50PM +0200, Yanxin Z. wrote:
> > Hello Igor Sysoev,
> > Thank you for your help.
> > 
> > You suggestion is great, but my problem is a little bit complicated.
> > 
> > The image name is flexible, can be from 0.jpg to 23456.jpg, we organize 
> > the image files into folders, whose name is the number divide by 1000. 
> > For example, the folder 100 saves the file 100000.jpg to 100999.jpg.
> > 
> > Both the folder and file name are flexible in my case.
> > 
> > Do you have any suggestion on it?
> 
> location /img/ {
> 
>     location ~ ^/img/(\d)(\.jpg)$ {
>         alias   /path/to/img/0$1/$1$2;
>     }
> 
>     location ~ ^/img/(\d\d)(\d*\.jpg)$ {
>         alias   /path/to/img/$1/$1$2;
>     }
> 
>     return 404;
> }

As far as I understand original question, the following should 
work (same as above, but fixed to use division by 1000 instead of 
two first digits):

location /img/ {

    location ~ ^/img/(\d+)(\d\d\d\.jpg)$ {
        alias /path/to/img/$1/$1$2;
    }

    location ~ ^/img/(\d+\.jpg)$ {
        alias /path/to/img/0/$1;
    }

    return 404;
}

Maxim Dounin


From mdounin at mdounin.ru  Mon May  2 23:26:17 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 2 May 2011 23:26:17 +0400
Subject: Shared nginx configuration for multiple servers
In-Reply-To: <de91070ac3595c06a46807fd3c68a3bb.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTimxeg3-hW=9x0K7-Q+RdC2D5fBJAQ@mail.gmail.com>
	<de91070ac3595c06a46807fd3c68a3bb.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110502192617.GG42265@mdounin.ru>

Hello!

On Mon, May 02, 2011 at 03:04:48PM -0400, dbanks wrote:

> I'm facing a related issue, although my attempt at implementation is
> different from the OP.
> 
> ip.conf:
> set $server_ip 10.11.1.50;
> 
> nginx.conf:
> ...
> server { 
>         include ip.conf;  #adds the $server_ip variable which is
> machine-specific
>         listen  $server_ip:80 default_server;
> ...
> }
> 
> This fails with:  nginx: [emerg] host not found in "$server_ip:80" of
> the "listen" directive in /opt/nginx/conf/nginx.conf:48
> 
> What is the best way to include a hard-coded IP address from another
> file in order to allow nginx.conf to be universal across deployed
> machines?

Variable are evaluated at runtime during request processing, you 
can't use them to configure listening sockets.

Either place full listen directive in included file or use some 
external scripting to produce actual config.

Maxim Dounin


From lists at ruby-forum.com  Mon May  2 23:29:55 2011
From: lists at ruby-forum.com (Yanxin Z.)
Date: Mon, 02 May 2011 21:29:55 +0200
Subject: nginx rewrite
In-Reply-To: <721e551f14c3665cad91466a13a63001@ruby-forum.com>
References: <721e551f14c3665cad91466a13a63001@ruby-forum.com>
Message-ID: <3f4dd13fbd8403d956188afc8bc69fd5@ruby-forum.com>

Thank you Maxim and Igor.
It's working for me.
Yanxin

-- 
Posted via http://www.ruby-forum.com/.


From mdounin at mdounin.ru  Tue May  3 01:04:56 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 3 May 2011 01:04:56 +0400
Subject: Crash (double free or corruption) on trying to proxy to localhost
In-Reply-To: <BANLkTikAQ-FGHBUT4+eFiEQ2stY88f10ow@mail.gmail.com>
References: <BANLkTikAQ-FGHBUT4+eFiEQ2stY88f10ow@mail.gmail.com>
Message-ID: <20110502210456.GH42265@mdounin.ru>

Hello!

On Sun, May 01, 2011 at 05:54:17PM -0700, Stephen Weeks wrote:

> I've currently got a pool of systems running nginx proxying to a pool of
> systems running apache.  I'm trying to move to running nginx locally on the
> apache hosts instead of separate systems to avoid the extra network hop,
> make more-efficient use of resources, and enable some future development
> (including migrating to running our application on nginx via fastcgi instead
> of apache, ideally).  we've currently got some significant architecture
> built up around apache, so converting right now is uncomfortable.
> 
> Ideally, I'd like nginx to just serve from localhost, but fail over to the
> rest of the pool when localhost in unavailable, so in my upstream I have
> every server except for localhost set as 'backup'.  I'm otherwise running
> identical configurations of apache and nginx on a single system together as
> used in the rest of the two pools.  This works exactly as expected, except
> that I get a few crashes of nginx workers every minute.  This only happens
> when proxying to the local system.  If I proxy anywhere else, it works
> fine.  Other proxies can serve from this system without trouble.  I see this
> same behaviour on other hosts when I build them the same way, so it's not an
> error with the host.  I see this crash on 0.7.65, 0.8.54, and 1.0.0, running
> on Ubuntu 10.04 LTS.  I see this crash whether I'm connecting to 127.0.0.1
> or the host's local IP.  I see this crash whether I'm listening on *:80 or
> <public ip>:80.  I see this crash whether I'm connecting to :80 or running
> apache on a different port and connecting to :81.  I see this crash whether
> I'm running ubuntu's "nginx-light" configuration, or their "nginx-full"
> configuration.  I see no errors logged from apache.
> 
> 1) I'd really love to make this work, so if there's anything else I can try,
> any additional debugging information I can give, I'd appreciate it.
> 2) Nginx has been very useful to me so far, so I thought you'd appreciate a
> bug report.
> 
> Posted on github, I have a problem description, section of a debug log, my
> (slightly edited: flattened includes and stripped an IP) nginx.conf, a gdb
> backtrace, and some additional information I was asked for when looking for
> help on IRC.  This is everything I've been able to come up with that sounds
> plausibly relevant.
> 
> https://gist.github.com/1574dbaf3a3dcda920a2
> 
> Any help?

Could you please provide:

1. nginx -V output

2. Full debug log for '*60' connection (the one which triggered 
abort in glibc), the one you provided contatins only last part of 
the connection in question.  Running grep -F ' 27772#0: *60 ' on 
original debug log should produce something useable.

Maxim Dounin


From ionathan at gmail.com  Tue May  3 02:00:18 2011
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Mon, 2 May 2011 19:00:18 -0300
Subject: reloading conf
Message-ID: <BANLkTike1cR3dMziW=vBHszMD8Zs2Mdk0w@mail.gmail.com>

Hi! How reloading my conf can affect my running nginx instances that are
under load?
Just made a small test and I saw that some requests to the upstream are
being lost. I am reloading the same configuration I had, didn't change
anything.

Is there any best practice on how to do it? how often? if I should avoid
something in particular? etc...

Thanks for your help!

Jonathan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110502/b10e8481/attachment.html>

From nginx-forum at nginx.us  Tue May  3 02:35:50 2011
From: nginx-forum at nginx.us (chiefops)
Date: Mon, 02 May 2011 18:35:50 -0400
Subject: domain rewrite/redirect
Message-ID: <9846569c04e02f90d5435cd81c105ee3.NginxMailingListEnglish@forum.nginx.org>

I have two different domains. One is a shortened url for referrals.   I
want to be able redirect the host name from the short url with
parameters to the actual url.  For example:

abcd.com/123456

rewrite/redirect to

efgh.com/referrals/123456

Initially I used a CNAME setup for adbcd.com to rewrite to efgh.com but
that doesn't rewrite the domain name or url.

Here's a short version of my setup with the rewrite rules I'm trying to
use:

http {
   server {
      listen 80;
      server_name abcd.com efgh.com;
      rewrite ^(.*) https://$host$1 permanent;
   }

   server {
      listen 443;
      server_name abcd.com efgh.com;
      if ($host ~* ^(efgh\.com)$ ) {
         rewrite ^/(.*)$ https://abcd.com/referrals/$ redirect;
      }
   }
}

The url on the user's browser after entering https://efgh.com/123456
should appear as https://abcd.com/referrals/123456.  Currently, it's
just rewriting.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195075,195075#msg-195075



From nginx-forum at nginx.us  Tue May  3 03:05:04 2011
From: nginx-forum at nginx.us (chiefops)
Date: Mon, 02 May 2011 19:05:04 -0400
Subject: domain rewrite/redirect
In-Reply-To: <9846569c04e02f90d5435cd81c105ee3.NginxMailingListEnglish@forum.nginx.org>
References: <9846569c04e02f90d5435cd81c105ee3.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <f9b1e17547c40759d56a6ad993ef67f7.NginxMailingListEnglish@forum.nginx.org>

I think I'm getting a bit further.  

I modified this a bit and its redirecting but not carrying the
parameters. Here's the updated config:

http {
   server {
      listen 80;
      server_name abcd.com;
      rewrite ^(.*) https://$host$1 permanent;
   }
   server {
      listen 80;
      server_name efgh.com;
      rewrite ^(.*) https://abcd.com/referrals/$1 redirect;
   }
   server {
      listen 443;
      server_name abcd.com;
   }
   server {
      listen 443;
      server efgh.com
      rewrite ^(.*) https://abcd.com/referrals/$1 permanent;
   }
}      
   }
}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195075,195084#msg-195084



From sweeks at imvu.com  Tue May  3 03:45:13 2011
From: sweeks at imvu.com (Stephen Weeks)
Date: Mon, 2 May 2011 16:45:13 -0700
Subject: Crash (double free or corruption) on trying to proxy to localhost
In-Reply-To: <20110502210456.GH42265@mdounin.ru>
References: <BANLkTikAQ-FGHBUT4+eFiEQ2stY88f10ow@mail.gmail.com>
	<20110502210456.GH42265@mdounin.ru>
Message-ID: <BANLkTinfMYbKy4h8vBxWwQrHrm+BtvnurA@mail.gmail.com>

Sure!  I've added it to the post on github.  I've slightly edited the log
out of paranoia (replacing a customer ID and auth key).

I can confirm now that I ran this proxy with the upstreamfair module
overnight and it didn't crash at all.  I can't get it to preferentially
serve to localhost, though, as it doesn't support 'backup' as a server
attribute, and doesn't seem to really use the weights, so that's suboptimal.

Anything else I can add to help troubleshooting this?  Anything you'd like
from the core dump?

On Mon, May 2, 2011 at 2:04 PM, Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
>
> On Sun, May 01, 2011 at 05:54:17PM -0700, Stephen Weeks wrote:
>
> > I've currently got a pool of systems running nginx proxying to a pool of
> > systems running apache.  I'm trying to move to running nginx locally on
> the
> > apache hosts instead of separate systems to avoid the extra network hop,
> > make more-efficient use of resources, and enable some future development
> > (including migrating to running our application on nginx via fastcgi
> instead
> > of apache, ideally).  we've currently got some significant architecture
> > built up around apache, so converting right now is uncomfortable.
> >
> > Ideally, I'd like nginx to just serve from localhost, but fail over to
> the
> > rest of the pool when localhost in unavailable, so in my upstream I have
> > every server except for localhost set as 'backup'.  I'm otherwise running
> > identical configurations of apache and nginx on a single system together
> as
> > used in the rest of the two pools.  This works exactly as expected,
> except
> > that I get a few crashes of nginx workers every minute.  This only
> happens
> > when proxying to the local system.  If I proxy anywhere else, it works
> > fine.  Other proxies can serve from this system without trouble.  I see
> this
> > same behaviour on other hosts when I build them the same way, so it's not
> an
> > error with the host.  I see this crash on 0.7.65, 0.8.54, and 1.0.0,
> running
> > on Ubuntu 10.04 LTS.  I see this crash whether I'm connecting to
> 127.0.0.1
> > or the host's local IP.  I see this crash whether I'm listening on *:80
> or
> > <public ip>:80.  I see this crash whether I'm connecting to :80 or
> running
> > apache on a different port and connecting to :81.  I see this crash
> whether
> > I'm running ubuntu's "nginx-light" configuration, or their "nginx-full"
> > configuration.  I see no errors logged from apache.
> >
> > 1) I'd really love to make this work, so if there's anything else I can
> try,
> > any additional debugging information I can give, I'd appreciate it.
> > 2) Nginx has been very useful to me so far, so I thought you'd appreciate
> a
> > bug report.
> >
> > Posted on github, I have a problem description, section of a debug log,
> my
> > (slightly edited: flattened includes and stripped an IP) nginx.conf, a
> gdb
> > backtrace, and some additional information I was asked for when looking
> for
> > help on IRC.  This is everything I've been able to come up with that
> sounds
> > plausibly relevant.
> >
> > https://gist.github.com/1574dbaf3a3dcda920a2
> >
> > Any help?
>
> Could you please provide:
>
> 1. nginx -V output
>
> 2. Full debug log for '*60' connection (the one which triggered
> abort in glibc), the one you provided contatins only last part of
> the connection in question.  Running grep -F ' 27772#0: *60 ' on
> original debug log should produce something useable.
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110502/dee6ab86/attachment-0001.html>

From nginx-forum at nginx.us  Tue May  3 03:55:11 2011
From: nginx-forum at nginx.us (chiefops)
Date: Mon, 02 May 2011 19:55:11 -0400
Subject: domain rewrite/redirect
In-Reply-To: <9846569c04e02f90d5435cd81c105ee3.NginxMailingListEnglish@forum.nginx.org>
References: <9846569c04e02f90d5435cd81c105ee3.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <7b57cba3f10768842326d51ac87ced5f.NginxMailingListEnglish@forum.nginx.org>

Nevermind, got if fixed on my end. Simple, just had to reread the wiki:

 rewrite ^ https://abcd.com/referrals$request_uri? permanent;

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195075,195093#msg-195093



From mdounin at mdounin.ru  Tue May  3 04:27:51 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 3 May 2011 04:27:51 +0400
Subject: reloading conf
In-Reply-To: <BANLkTike1cR3dMziW=vBHszMD8Zs2Mdk0w@mail.gmail.com>
References: <BANLkTike1cR3dMziW=vBHszMD8Zs2Mdk0w@mail.gmail.com>
Message-ID: <20110503002750.GJ42265@mdounin.ru>

Hello!

On Mon, May 02, 2011 at 07:00:18PM -0300, Jonathan Leibiusky wrote:

> Hi! How reloading my conf can affect my running nginx instances that are
> under load?

While reloading nginx master process parses new configuration, 
starts new worker processes and instructs old worker processes to 
gracefully terminate (i.e. finish processing of all requests and 
then exit).

This shouldn't affect anything but may require additional 
resources (mostly memory).  If you don't have enough memory reload 
may be problematic operation, and it's generally good idea to make 
sure your running nginx doesn't consume more than half of your 
server memory under normal load (to make sure it will be possible 
to start additional worker processes on reload).

> Just made a small test and I saw that some requests to the upstream are
> being lost. I am reloading the same configuration I had, didn't change
> anything.

It's strange you see something being lost.  Are you sure you 
actually did reload (i.e. sent SIGHUP to nginx master process)?

> Is there any best practice on how to do it? how often? if I should avoid
> something in particular? etc...

General practice is to reload config when you've changed it.

If you are doing reloads often - you may want to make sure 
previous reload was completed (i.e. old workers terminated) before 
doing another one (there is nothing fatal if it's not, but see 
above about resources).  Please note that with some use cases 
(long-running requests, downloads and so on) reload may take 
several hours or even more.

Maxim Dounin


From ionathan at gmail.com  Tue May  3 05:02:21 2011
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Mon, 2 May 2011 22:02:21 -0300
Subject: reloading conf
In-Reply-To: <20110503002750.GJ42265@mdounin.ru>
References: <BANLkTike1cR3dMziW=vBHszMD8Zs2Mdk0w@mail.gmail.com>
	<20110503002750.GJ42265@mdounin.ru>
Message-ID: <BANLkTikxgLKzYUODkEf6URzkPAppgWceBA@mail.gmail.com>

Thanks! I will check it again. In general my nginx instances have very
low cpu and memory usage, way beneath 50%, probably around 10% or even
less.
And request in our app are avg 200 or 300 ms, so no long running requests.
We are serving around 15k req/s on 5 nginx instances
(instance=machine), so maybe we have something wrong in the conf or
something.
Anyway, now I understand how it works, so it shouldn't affect anything.

Thanks again!

Jonathan

On 5/2/11, Maxim Dounin <mdounin at mdounin.ru> wrote:
> Hello!
>
> On Mon, May 02, 2011 at 07:00:18PM -0300, Jonathan Leibiusky wrote:
>
>> Hi! How reloading my conf can affect my running nginx instances that are
>> under load?
>
> While reloading nginx master process parses new configuration,
> starts new worker processes and instructs old worker processes to
> gracefully terminate (i.e. finish processing of all requests and
> then exit).
>
> This shouldn't affect anything but may require additional
> resources (mostly memory).  If you don't have enough memory reload
> may be problematic operation, and it's generally good idea to make
> sure your running nginx doesn't consume more than half of your
> server memory under normal load (to make sure it will be possible
> to start additional worker processes on reload).
>
>> Just made a small test and I saw that some requests to the upstream are
>> being lost. I am reloading the same configuration I had, didn't change
>> anything.
>
> It's strange you see something being lost.  Are you sure you
> actually did reload (i.e. sent SIGHUP to nginx master process)?
>
>> Is there any best practice on how to do it? how often? if I should avoid
>> something in particular? etc...
>
> General practice is to reload config when you've changed it.
>
> If you are doing reloads often - you may want to make sure
> previous reload was completed (i.e. old workers terminated) before
> doing another one (there is nothing fatal if it's not, but see
> above about resources).  Please note that with some use cases
> (long-running requests, downloads and so on) reload may take
> several hours or even more.
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>


From lists at ruby-forum.com  Tue May  3 06:03:29 2011
From: lists at ruby-forum.com (Mike P.)
Date: Tue, 03 May 2011 04:03:29 +0200
Subject: 502 errors, upstream sent too big header
In-Reply-To: <BANLkTi=CkxwWv5aW8F+Rxn6DCanaebLnhQ@mail.gmail.com>
References: <BANLkTi=CkxwWv5aW8F+Rxn6DCanaebLnhQ@mail.gmail.com>
Message-ID: <d64f79579160c2cd575b3b2a31feb49c@ruby-forum.com>

According to passenger's documentation, it doesn't appear that a
directive exists to adjust any header size limitations.

Also, Phusion's CTO doesn't mention any possible solution here regarding 
the same error: http://goo.gl/ZPXvO

If there is no such adjustment that can be made via passenger to
alleviate this error, is there also no possible solution within Nginx?

Thanks!

-- 
Posted via http://www.ruby-forum.com/.


From igor at sysoev.ru  Tue May  3 08:19:01 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 3 May 2011 08:19:01 +0400
Subject: 502 errors, upstream sent too big header
In-Reply-To: <BANLkTi=CkxwWv5aW8F+Rxn6DCanaebLnhQ@mail.gmail.com>
References: <BANLkTi=CkxwWv5aW8F+Rxn6DCanaebLnhQ@mail.gmail.com>
Message-ID: <20110503041901.GA84770@sysoev.ru>

On Mon, Apr 04, 2011 at 04:52:22PM -0400, Travis Derouin wrote:
> We seem to be getting sporadic 502 errors, mostly when users are logging in.
> We run a Mediawiki based website, are running nginx 0.9.5, php-fpm 5.3.5.
> 
> The error:
> 
> 2011/04/04 10:41:34 [error] 30517#0: *2143711 upstream sent too big header
> while reading response header from upstream, client: 192.168.100.5, server:
> _, request: "POST
> /index.php?title=Special:Userlogin&action=submitlogin&type=login&returnto=Main-Page
> HTTP/1.1", upstream: "fastcgi://127.0.0.1:9000", host: "www.wikihow.com",
> referrer: "http://www.wikihow.com/Special:Userlogin?returnto=Main-Page"
> 
> Our nginx.conf proxy buffer settings:
> 
>     proxy_buffers 16 16k;
>     proxy_buffer_size 32k;
> 
> We've tried increasing these values from the defaults a few times, and it
> doesn't seem to fix the issue.
> 
> Do you have any suggestions on how to fix this?

fastcgi_buffers  16 16k;
fastcgi_buffer_size  32k;

since you use fastcgi, but not proxy.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Tue May  3 10:09:45 2011
From: nginx-forum at nginx.us (moorthi)
Date: Tue, 03 May 2011 02:09:45 -0400
Subject: nginx imaps auth_http cyrus
In-Reply-To: <8599d7049c20a5f08721167b39775a44.NginxMailingListEnglish@forum.nginx.org>
References: <4D73FB1E.3090104@alokat.org>
	<3f29c76ca3fec62ea3f4f3423b70de4e.NginxMailingListEnglish@forum.nginx.org>
	<8599d7049c20a5f08721167b39775a44.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <656fc93354f0009fb1ef5831e7369a2a.NginxMailingListEnglish@forum.nginx.org>

auth_http script is working fine if i use pop/imap ports 110/143, if i
use 993/995 it's not working

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,180902,195125#msg-195125



From cedric.jeanneret at camptocamp.com  Tue May  3 11:44:48 2011
From: cedric.jeanneret at camptocamp.com (Cedric Jeanneret)
Date: Tue, 3 May 2011 09:44:48 +0200
Subject: domain rewrite/redirect
In-Reply-To: <7b57cba3f10768842326d51ac87ced5f.NginxMailingListEnglish@forum.nginx.org>
References: <9846569c04e02f90d5435cd81c105ee3.NginxMailingListEnglish@forum.nginx.org>
	<7b57cba3f10768842326d51ac87ced5f.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110503094448.27d397f8@cholatse.wrk.lsn.camptocamp.com>

On Mon, 02 May 2011 19:55:11 -0400
"chiefops" <nginx-forum at nginx.us> wrote:

> Nevermind, got if fixed on my end. Simple, just had to reread the wiki:
> 
>  rewrite ^ https://abcd.com/referrals$request_uri? permanent;
> 
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195075,195093#msg-195093
> 
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx

Hello,

You may also use HSTS headers to ensure your users use SSL instead of non-SSL. This won't replace the redirect, but the compatible browser will know on the next visit that it should use the SSL server only. It won't request the non-SSL part.

More about HSTS:
https://secure.wikimedia.org/wikipedia/en/wiki/HTTP_Strict_Transport_Security

Cheers,

C.

-- 
C?dric Jeanneret                 |  System Administrator
021 619 10 32                    |  Camptocamp SA
cedric.jeanneret at camptocamp.com  |  PSE-A / EPFL
www.camptocamp.com               | www.github.com/camptocamp
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://nginx.org/pipermail/nginx/attachments/20110503/db2bbc86/attachment.pgp>

From mdounin at mdounin.ru  Tue May  3 11:45:07 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 3 May 2011 11:45:07 +0400
Subject: 502 errors, upstream sent too big header
In-Reply-To: <d64f79579160c2cd575b3b2a31feb49c@ruby-forum.com>
References: <BANLkTi=CkxwWv5aW8F+Rxn6DCanaebLnhQ@mail.gmail.com>
	<d64f79579160c2cd575b3b2a31feb49c@ruby-forum.com>
Message-ID: <20110503074507.GL42265@mdounin.ru>

Hello!

On Tue, May 03, 2011 at 04:03:29AM +0200, Mike P. wrote:

> According to passenger's documentation, it doesn't appear that a
> directive exists to adjust any header size limitations.
> 
> Also, Phusion's CTO doesn't mention any possible solution here regarding 
> the same error: http://goo.gl/ZPXvO

Ah, indeed: looking though code suggests they unconditionally set 
buffer size to pagesize and doesn't allow it to be changed.

> If there is no such adjustment that can be made via passenger to
> alleviate this error, is there also no possible solution within Nginx?

You may consider using another module to talk to backends, 
standard proxy or fastcgi may be a good solution.

Maxim Dounin


From igor at sysoev.ru  Tue May  3 13:10:55 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 3 May 2011 13:10:55 +0400
Subject: Using map and proxy_pass
In-Reply-To: <201104301252.43466.markus.jelsma@openindex.io>
References: <201104291748.48357.markus.jelsma@openindex.io>
	<f975fd982b56ba3953d51b7e1892d895.NginxMailingListEnglish@forum.nginx.org>
	<201104301252.43466.markus.jelsma@openindex.io>
Message-ID: <20110503091055.GE1707@sysoev.ru>

On Sat, Apr 30, 2011 at 12:52:42PM +0200, Markus Jelsma wrote:
> Too many. I need to generate the mappings in an external program. I'd rather 
> not generate as many different locations.

Why ? locations without regex will run fast.


-- 
Igor Sysoev


From mscherer82 at gmail.com  Tue May  3 15:01:53 2011
From: mscherer82 at gmail.com (mscherer82 at gmail.com)
Date: Tue, 3 May 2011 13:01:53 +0200
Subject: multiple connections to a fastcgi server
Message-ID: <BANLkTinT=PHDz4banYL8isdrD3py3DEOtA@mail.gmail.com>

Hello,

I want to build a fast cgi server application which can handle
multiple simoultanous calls. I found the threaded.c (or threaded.pl)
example in the fast cgi sdk.
(http://www.fastcgi.com/devkit/examples/threaded.c)

But when running with my current configuration, the webserver only
opens one connection at time and the second requests is waiting until
the first one is finished. I don't want to use multiplexing. It could
be done through multiple connections.


default:

server {
...
       location ~ \.c$ {
         include /etc/nginx/fastcgi_params; #or whatever you named it
       #  fastcgi_pass unix:/tmp/c.socket;
         fastcgi_pass 127.0.0.1:9000;
       }
}

Is it possible to change this behaviour?

Thanks
Marc


From markus.jelsma at openindex.io  Tue May  3 15:02:53 2011
From: markus.jelsma at openindex.io (Markus Jelsma)
Date: Tue, 3 May 2011 13:02:53 +0200
Subject: Using map and proxy_pass
In-Reply-To: <20110503091055.GE1707@sysoev.ru>
References: <201104291748.48357.markus.jelsma@openindex.io>
	<201104301252.43466.markus.jelsma@openindex.io>
	<20110503091055.GE1707@sysoev.ru>
Message-ID: <201105031302.53844.markus.jelsma@openindex.io>

Thanks Francis and Igor. Is there are penalty for having a huge configuration 
because of a very large number of locations?

On Tuesday 03 May 2011 11:10:55 Igor Sysoev wrote:
> On Sat, Apr 30, 2011 at 12:52:42PM +0200, Markus Jelsma wrote:
> > Too many. I need to generate the mappings in an external program. I'd
> > rather not generate as many different locations.
> 
> Why ? locations without regex will run fast.

-- 
Markus Jelsma - CTO - Openindex
http://www.linkedin.com/in/markus17
050-8536620 / 06-50258350


From mdounin at mdounin.ru  Tue May  3 15:46:42 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 3 May 2011 15:46:42 +0400
Subject: Crash (double free or corruption) on trying to proxy to localhost
In-Reply-To: <BANLkTinfMYbKy4h8vBxWwQrHrm+BtvnurA@mail.gmail.com>
References: <BANLkTikAQ-FGHBUT4+eFiEQ2stY88f10ow@mail.gmail.com>
	<20110502210456.GH42265@mdounin.ru>
	<BANLkTinfMYbKy4h8vBxWwQrHrm+BtvnurA@mail.gmail.com>
Message-ID: <20110503114642.GN42265@mdounin.ru>

Hello!

On Mon, May 02, 2011 at 04:45:13PM -0700, Stephen Weeks wrote:

> Sure!  I've added it to the post on github.  I've slightly edited the log
> out of paranoia (replacing a customer ID and auth key).
> 
> I can confirm now that I ran this proxy with the upstreamfair module
> overnight and it didn't crash at all.  I can't get it to preferentially
> serve to localhost, though, as it doesn't support 'backup' as a server
> attribute, and doesn't seem to really use the weights, so that's suboptimal.
> 
> Anything else I can add to help troubleshooting this?  Anything you'd like
> from the core dump?

Ok, thank you, it looks like I see the problem.

Allocation for "tried" flags doesn't take into account number of 
backup servers, and if there are more backup servers than normal 
ones (and backup servers are in fact used) - this may cause memory 
corruption.

Please try the attached patch.

Maxim Dounin

> 
> On Mon, May 2, 2011 at 2:04 PM, Maxim Dounin <mdounin at mdounin.ru> wrote:
> 
> > Hello!
> >
> > On Sun, May 01, 2011 at 05:54:17PM -0700, Stephen Weeks wrote:
> >
> > > I've currently got a pool of systems running nginx proxying to a pool of
> > > systems running apache.  I'm trying to move to running nginx locally on
> > the
> > > apache hosts instead of separate systems to avoid the extra network hop,
> > > make more-efficient use of resources, and enable some future development
> > > (including migrating to running our application on nginx via fastcgi
> > instead
> > > of apache, ideally).  we've currently got some significant architecture
> > > built up around apache, so converting right now is uncomfortable.
> > >
> > > Ideally, I'd like nginx to just serve from localhost, but fail over to
> > the
> > > rest of the pool when localhost in unavailable, so in my upstream I have
> > > every server except for localhost set as 'backup'.  I'm otherwise running
> > > identical configurations of apache and nginx on a single system together
> > as
> > > used in the rest of the two pools.  This works exactly as expected,
> > except
> > > that I get a few crashes of nginx workers every minute.  This only
> > happens
> > > when proxying to the local system.  If I proxy anywhere else, it works
> > > fine.  Other proxies can serve from this system without trouble.  I see
> > this
> > > same behaviour on other hosts when I build them the same way, so it's not
> > an
> > > error with the host.  I see this crash on 0.7.65, 0.8.54, and 1.0.0,
> > running
> > > on Ubuntu 10.04 LTS.  I see this crash whether I'm connecting to
> > 127.0.0.1
> > > or the host's local IP.  I see this crash whether I'm listening on *:80
> > or
> > > <public ip>:80.  I see this crash whether I'm connecting to :80 or
> > running
> > > apache on a different port and connecting to :81.  I see this crash
> > whether
> > > I'm running ubuntu's "nginx-light" configuration, or their "nginx-full"
> > > configuration.  I see no errors logged from apache.
> > >
> > > 1) I'd really love to make this work, so if there's anything else I can
> > try,
> > > any additional debugging information I can give, I'd appreciate it.
> > > 2) Nginx has been very useful to me so far, so I thought you'd appreciate
> > a
> > > bug report.
> > >
> > > Posted on github, I have a problem description, section of a debug log,
> > my
> > > (slightly edited: flattened includes and stripped an IP) nginx.conf, a
> > gdb
> > > backtrace, and some additional information I was asked for when looking
> > for
> > > help on IRC.  This is everything I've been able to come up with that
> > sounds
> > > plausibly relevant.
> > >
> > > https://gist.github.com/1574dbaf3a3dcda920a2
> > >
> > > Any help?
> >
> > Could you please provide:
> >
> > 1. nginx -V output
> >
> > 2. Full debug log for '*60' connection (the one which triggered
> > abort in glibc), the one you provided contatins only last part of
> > the connection in question.  Running grep -F ' 27772#0: *60 ' on
> > original debug log should produce something useable.
> >
> > Maxim Dounin
> >
> > _______________________________________________
> > nginx mailing list
> > nginx at nginx.org
> > http://nginx.org/mailman/listinfo/nginx
> >

> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx

-------------- next part --------------
# HG changeset patch
# User Maxim Dounin <mdounin at mdounin.ru>
# Date 1304422854 -14400
# Node ID b7826a837aaf484462db58d64ec0d060ba8b92e5
# Parent  00d13b6d4ebd225f94a2e2a3afa7dbd3ddfe4ed7
Upstream: properly allocate memory for tried flags.

Previous allocation only took into account number of non-backup servers, and
this caused memory corruption with many backup servers.

See report here:

http://nginx.org/pipermail/nginx/2011-May/026531.html

diff --git a/src/http/ngx_http_upstream_round_robin.c b/src/http/ngx_http_upstream_round_robin.c
--- a/src/http/ngx_http_upstream_round_robin.c
+++ b/src/http/ngx_http_upstream_round_robin.c
@@ -219,13 +219,18 @@ ngx_http_upstream_init_round_robin_peer(
     rrp->peers = us->peer.data;
     rrp->current = 0;
 
-    if (rrp->peers->number <= 8 * sizeof(uintptr_t)) {
+    n = rrp->peers->number;
+
+    if (rrp->peers->next && rrp->peers->next->number > n) {
+        n = rrp->peers->next->number;
+    }
+
+    if (n <= 8 * sizeof(uintptr_t)) {
         rrp->tried = &rrp->data;
         rrp->data = 0;
 
     } else {
-        n = (rrp->peers->number + (8 * sizeof(uintptr_t) - 1))
-                / (8 * sizeof(uintptr_t));
+        n = (n + (8 * sizeof(uintptr_t) - 1)) / (8 * sizeof(uintptr_t));
 
         rrp->tried = ngx_pcalloc(r->pool, n * sizeof(uintptr_t));
         if (rrp->tried == NULL) {

From sweeks at imvu.com  Tue May  3 16:00:03 2011
From: sweeks at imvu.com (Stephen Weeks)
Date: Tue, 3 May 2011 05:00:03 -0700
Subject: Crash (double free or corruption) on trying to proxy to localhost
In-Reply-To: <20110503114642.GN42265@mdounin.ru>
References: <BANLkTikAQ-FGHBUT4+eFiEQ2stY88f10ow@mail.gmail.com>
	<20110502210456.GH42265@mdounin.ru>
	<BANLkTinfMYbKy4h8vBxWwQrHrm+BtvnurA@mail.gmail.com>
	<20110503114642.GN42265@mdounin.ru>
Message-ID: <BANLkTinevo=ZUtjpHJNaMGUsiAFx6PXpuw@mail.gmail.com>

I should be able to try this out tonight (>12 hours from now).  I'll let you
know how it works.  Thanks for looking into this for me.

On Tue, May 3, 2011 at 4:46 AM, Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
>
> On Mon, May 02, 2011 at 04:45:13PM -0700, Stephen Weeks wrote:
>
> > Sure!  I've added it to the post on github.  I've slightly edited the log
> > out of paranoia (replacing a customer ID and auth key).
> >
> > I can confirm now that I ran this proxy with the upstreamfair module
> > overnight and it didn't crash at all.  I can't get it to preferentially
> > serve to localhost, though, as it doesn't support 'backup' as a server
> > attribute, and doesn't seem to really use the weights, so that's
> suboptimal.
> >
> > Anything else I can add to help troubleshooting this?  Anything you'd
> like
> > from the core dump?
>
> Ok, thank you, it looks like I see the problem.
>
> Allocation for "tried" flags doesn't take into account number of
> backup servers, and if there are more backup servers than normal
> ones (and backup servers are in fact used) - this may cause memory
> corruption.
>
> Please try the attached patch.
>
> Maxim Dounin
>
> >
> > On Mon, May 2, 2011 at 2:04 PM, Maxim Dounin <mdounin at mdounin.ru> wrote:
> >
> > > Hello!
> > >
> > > On Sun, May 01, 2011 at 05:54:17PM -0700, Stephen Weeks wrote:
> > >
> > > > I've currently got a pool of systems running nginx proxying to a pool
> of
> > > > systems running apache.  I'm trying to move to running nginx locally
> on
> > > the
> > > > apache hosts instead of separate systems to avoid the extra network
> hop,
> > > > make more-efficient use of resources, and enable some future
> development
> > > > (including migrating to running our application on nginx via fastcgi
> > > instead
> > > > of apache, ideally).  we've currently got some significant
> architecture
> > > > built up around apache, so converting right now is uncomfortable.
> > > >
> > > > Ideally, I'd like nginx to just serve from localhost, but fail over
> to
> > > the
> > > > rest of the pool when localhost in unavailable, so in my upstream I
> have
> > > > every server except for localhost set as 'backup'.  I'm otherwise
> running
> > > > identical configurations of apache and nginx on a single system
> together
> > > as
> > > > used in the rest of the two pools.  This works exactly as expected,
> > > except
> > > > that I get a few crashes of nginx workers every minute.  This only
> > > happens
> > > > when proxying to the local system.  If I proxy anywhere else, it
> works
> > > > fine.  Other proxies can serve from this system without trouble.  I
> see
> > > this
> > > > same behaviour on other hosts when I build them the same way, so it's
> not
> > > an
> > > > error with the host.  I see this crash on 0.7.65, 0.8.54, and 1.0.0,
> > > running
> > > > on Ubuntu 10.04 LTS.  I see this crash whether I'm connecting to
> > > 127.0.0.1
> > > > or the host's local IP.  I see this crash whether I'm listening on
> *:80
> > > or
> > > > <public ip>:80.  I see this crash whether I'm connecting to :80 or
> > > running
> > > > apache on a different port and connecting to :81.  I see this crash
> > > whether
> > > > I'm running ubuntu's "nginx-light" configuration, or their
> "nginx-full"
> > > > configuration.  I see no errors logged from apache.
> > > >
> > > > 1) I'd really love to make this work, so if there's anything else I
> can
> > > try,
> > > > any additional debugging information I can give, I'd appreciate it.
> > > > 2) Nginx has been very useful to me so far, so I thought you'd
> appreciate
> > > a
> > > > bug report.
> > > >
> > > > Posted on github, I have a problem description, section of a debug
> log,
> > > my
> > > > (slightly edited: flattened includes and stripped an IP) nginx.conf,
> a
> > > gdb
> > > > backtrace, and some additional information I was asked for when
> looking
> > > for
> > > > help on IRC.  This is everything I've been able to come up with that
> > > sounds
> > > > plausibly relevant.
> > > >
> > > > https://gist.github.com/1574dbaf3a3dcda920a2
> > > >
> > > > Any help?
> > >
> > > Could you please provide:
> > >
> > > 1. nginx -V output
> > >
> > > 2. Full debug log for '*60' connection (the one which triggered
> > > abort in glibc), the one you provided contatins only last part of
> > > the connection in question.  Running grep -F ' 27772#0: *60 ' on
> > > original debug log should produce something useable.
> > >
> > > Maxim Dounin
> > >
> > > _______________________________________________
> > > nginx mailing list
> > > nginx at nginx.org
> > > http://nginx.org/mailman/listinfo/nginx
> > >
>
> > _______________________________________________
> > nginx mailing list
> > nginx at nginx.org
> > http://nginx.org/mailman/listinfo/nginx
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110503/1f500fa2/attachment.html>

From igor at sysoev.ru  Tue May  3 16:06:21 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 3 May 2011 16:06:21 +0400
Subject: Using map and proxy_pass
In-Reply-To: <201105031302.53844.markus.jelsma@openindex.io>
References: <201104291748.48357.markus.jelsma@openindex.io>
	<201104301252.43466.markus.jelsma@openindex.io>
	<20110503091055.GE1707@sysoev.ru>
	<201105031302.53844.markus.jelsma@openindex.io>
Message-ID: <20110503120621.GT1707@sysoev.ru>

On Tue, May 03, 2011 at 01:02:53PM +0200, Markus Jelsma wrote:
> Thanks Francis and Igor. Is there are penalty for having a huge configuration 
> because of a very large number of locations?

nginx uses some kind of binary tree to search locations, so penalty
should be small.


-- 
Igor Sysoev


> On Tuesday 03 May 2011 11:10:55 Igor Sysoev wrote:
> > On Sat, Apr 30, 2011 at 12:52:42PM +0200, Markus Jelsma wrote:
> > > Too many. I need to generate the mappings in an external program. I'd
> > > rather not generate as many different locations.
> > 
> > Why ? locations without regex will run fast.
> 
> -- 
> Markus Jelsma - CTO - Openindex
> http://www.linkedin.com/in/markus17
> 050-8536620 / 06-50258350


From igor at sysoev.ru  Tue May  3 16:15:58 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 3 May 2011 16:15:58 +0400
Subject: nginx-1.0.1
Message-ID: <20110503121558.GW1707@sysoev.ru>

Changes with nginx 1.0.1                                         03 May 2011

    *) Change: now the "split_clients" directive uses MurmurHash2 algorithm 
       because of better distribution.
       Thanks to Oleg Mamontov.

    *) Change: now long strings starting with zero are not considered as 
       false values.
       Thanks to Maxim Dounin.

    *) Change: now nginx uses a default listen backlog value 511 on Linux.

    *) Feature: the $upstream_... variables may be used in the SSI and perl 
       modules.

    *) Bugfix: now nginx limits better disk cache size.
       Thanks to Oleg Mamontov.

    *) Bugfix: a segmentation fault might occur while parsing incorrect 
       IPv4 address; the bug had appeared in 0.9.3.
       Thanks to Maxim Dounin.

    *) Bugfix: nginx could not be built by gcc 4.6 without --with-debug 
       option.

    *) Bugfix: nginx could not be built on Solaris 9 and earlier; the bug 
       had appeared in 0.9.3.
       Thanks to Dagobert Michelsen.

    *) Bugfix: $request_time variable had invalid values if subrequests 
       were used; the bug had appeared in 0.8.47.
       Thanks to Igor A. Valcov.


-- 
Igor Sysoev


From mdounin at mdounin.ru  Tue May  3 16:16:09 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 3 May 2011 16:16:09 +0400
Subject: multiple connections to a fastcgi server
In-Reply-To: <BANLkTinT=PHDz4banYL8isdrD3py3DEOtA@mail.gmail.com>
References: <BANLkTinT=PHDz4banYL8isdrD3py3DEOtA@mail.gmail.com>
Message-ID: <20110503121608.GQ42265@mdounin.ru>

Hello!

On Tue, May 03, 2011 at 01:01:53PM +0200, mscherer82 at gmail.com wrote:

> I want to build a fast cgi server application which can handle
> multiple simoultanous calls. I found the threaded.c (or threaded.pl)
> example in the fast cgi sdk.
> (http://www.fastcgi.com/devkit/examples/threaded.c)
> 
> But when running with my current configuration, the webserver only
> opens one connection at time and the second requests is waiting until
> the first one is finished. I don't want to use multiplexing. It could
> be done through multiple connections.

nginx will open as many connections to fastcgi backend as it has 
concurrent requests.

If you see only one connection at a time - you may want to check 
fastcgi app and testing tools you are using, probably 
serialization of requests happens there.

Maxim Dounin


From igor at sysoev.ru  Tue May  3 16:45:03 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 3 May 2011 16:45:03 +0400
Subject: How to get a 404 logged with try_files?
In-Reply-To: <eece71dda665fc6cf671ec665fe4db89.NginxMailingListEnglish@forum.nginx.org>
References: <eece71dda665fc6cf671ec665fe4db89.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110503124503.GZ1707@sysoev.ru>

On Mon, May 02, 2011 at 10:22:50AM -0400, monster wrote:
> Hi! Just started with Nginx after reading it reached 1.0.
> Congratulation!
> 
> I'm trying to get my test site to return a "missing image" icon when an
> image is missing, but have a 404 logged so I know I messed up something.
> I want to specify the images name without file extensions, in case I
> move some from PNG to JPG, or the other way around. My problem is that
> try_files doc says last path MUST exist. I would like last path to be
> $uri and cause a 404, not a 500. Tried using error_page to convert 500
> to 404, but did not have any effect. Thanks for any advice!
> 
> location ~ ^/img/[^\.]*$ {
>     try_files $uri.png $uri.jpg $uri.gif $uri;
>     error_page  500 =404	/img/missing;
> }
> 
> location /img/ {
>     error_page  404		/img/missing;
> }
> 
> location = /img/missing {
>     rewrite .* /img/icon/missing.png;
> }

location /img/ {
    error_page  404   /img/icon/missing.png;

    location ~ ^/img/[^\.]*$ {
        try_files $uri.png $uri.jpg $uri.gif $uri =404;
    }
}

location = /img/icon/missing.png {
    # nothing
}


-- 
Igor Sysoev


From nginx-forum at nginx.us  Tue May  3 18:24:54 2011
From: nginx-forum at nginx.us (darckos)
Date: Tue, 03 May 2011 10:24:54 -0400
Subject: Nginx Slowdown on Solaris
In-Reply-To: <ec21cadaf64b543c58a6436798533d95.NginxMailingListEnglish@forum.nginx.org>
References: <5b962f75c896719f311e1758cf349b78.NginxMailingListEnglish@forum.nginx.org>
	<b184a14137fee618c1527f88f433c9e0.NginxMailingListEnglish@forum.nginx.org>
	<28809a5d0d94cff2d6f534f448399488.NginxMailingListEnglish@forum.nginx.org>
	<ec21cadaf64b543c58a6436798533d95.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <42399eb9b544e9cb55d74d80ca601475.NginxMailingListEnglish@forum.nginx.org>

Hello, 

I just changed my config with option 3 and I feel like my nignx is
better now. 

http://img232.imageshack.us/img232/5369/nginx.png

My new config is : 

worker_processes  24;

events {
     worker_connections  2048;
     devpoll_events  1;
}

http {
    include       mime.types;
    default_type  application/octet-stream;

    output_buffers 2 256k;
    keepalive_timeout  15;

...
}

Thanks for your help

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,194233,195256#msg-195256



From piotr.sikora at frickle.com  Tue May  3 20:16:50 2011
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Tue, 3 May 2011 18:16:50 +0200
Subject: FRiCKLE Labs & yo.se pres. ngx_cache_purge
In-Reply-To: <8EB2CC25B0714F8B988CE183AE7BF6C9@Desktop>
References: <07864ADE7B1B46B19DA043DF075AD211@nightmare><F7FE465719AB4118934613F5E2056130@Desktop>
	<8EB2CC25B0714F8B988CE183AE7BF6C9@Desktop>
Message-ID: <2FF88DB4047D46DBB6B2CB508533EA85@Desktop>

Version 1.3 is now available at:
http://labs.frickle.com/nginx_ngx_cache_purge/

GitHub repository is available at:
http://github.com/FRiCKLE/ngx_cache_purge/

Changes:
2011-05-03    VERSION 1.3
    * Fix compatibility with nginx-1.0.1.
      Reported by Sergey A. Osokin and Markus Linnala.

2010-08-29
    * Fix compatibility with nginx-0.8.0 and versions older than
      nginx-0.7.60.

Best regards,
Piotr Sikora < piotr.sikora at frickle.com >



From piotr.sikora at frickle.com  Tue May  3 21:15:54 2011
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Tue, 3 May 2011 19:15:54 +0200
Subject: FRiCKLE Labs & C2Hosting pres. ngx_slowfs_cache
In-Reply-To: <9598C25E2C214D3AB981B7163E0C967A@Desktop>
References: <3F594E374AC64662B26C63FDACA6F09F@nightmare><5708F1AFE3324115B1591ACD4EE3775E@nightmare><491198975B004407BC79739160F49616@nightmare><AAA04C9D3429466A82E9041998877B5D@Desktop>
	<9598C25E2C214D3AB981B7163E0C967A@Desktop>
Message-ID: <59110DA6D3684E28A1471E0452991D85@Desktop>

Version 1.6 is now available at:
http://labs.frickle.com/nginx_ngx_slowfs_cache/

GitHub repository is available at:
http://github.com/FRiCKLE/ngx_slowfs_cache/

Changes:
2011-05-03    VERSION 1.6
    * Fix compatibility with nginx-1.0.1.
      Reported by Sergey A. Osokin.

2011-04-21
    * Stop doing sanity checks when module isn't enabled.
      Reported by Tobi Tamas.

Best regards,
Piotr Sikora < piotr.sikora at frickle.com >



From nginx-forum at nginx.us  Tue May  3 21:42:42 2011
From: nginx-forum at nginx.us (art.wu)
Date: Tue, 03 May 2011 13:42:42 -0400
Subject: no server response in https
Message-ID: <f2292582eeb12229fb3a95a01b37b61e.NginxMailingListEnglish@forum.nginx.org>

I'm running 0.8.53 on QNX Neutrino 6.5.0 with SSL on and the server
doesn't respond to the client. Instead it times out, as below. It does
work properly with SSL off. Can you provide any suggestions? Thanks.


2011/05/02 09:39:57 [debug] 2220074#0: *7 http check ssl handshake
2011/05/02 09:39:57 [debug] 2220074#0: *7 https ssl handshake: 0x16
2011/05/02 09:39:57 [debug] 2220074#0: *7 SSL_do_handshake: -1
2011/05/02 09:39:57 [debug] 2220074#0: *7 SSL_get_error: 2
2011/05/02 09:39:57 [debug] 2220074#0: *7 post event 080E284C
2011/05/02 09:39:57 [debug] 2220074#0: *7 delete posted event 080E284C
2011/05/02 09:39:57 [debug] 2220074#0: *7 SSL handshake handler: 0
2011/05/02 09:39:57 [debug] 2220074#0: *7 SSL_do_handshake: 1
2011/05/02 09:39:57 [debug] 2220074#0: *7 select del event fd:3 ev:5
2011/05/02 09:39:57 [debug] 2220074#0: *7 SSL: TLSv1, cipher:
"DHE-RSA-AES256-SHA SSLv3 Kx=DH Au=RSA Enc=AES(256) Mac=SHA1"
2011/05/02 09:39:57 [debug] 2220074#0: *7 http process request line
2011/05/02 09:39:57 [debug] 2220074#0: *7 SSL_read: -1
2011/05/02 09:39:57 [debug] 2220074#0: *7 SSL_get_error: 2
2011/05/02 09:39:57 [debug] 2220074#0: *7 select add event fd:3 ev:5
2011/05/02 09:39:57 [debug] 2220074#0: *7 post event 080E284C
2011/05/02 09:39:57 [debug] 2220074#0: *7 delete posted event 080E284C
2011/05/02 09:39:57 [debug] 2220074#0: *7 http process request line
2011/05/02 09:39:57 [debug] 2220074#0: *7 SSL_read: 132
2011/05/02 09:39:57 [debug] 2220074#0: *7 SSL_read: -1
2011/05/02 09:39:57 [debug] 2220074#0: *7 SSL_get_error: 2
2011/05/02 09:39:57 [debug] 2220074#0: *7 http request line: "GET /
HTTP/1.1"
2011/05/02 09:39:57 [debug] 2220074#0: *7 http uri: "/"
2011/05/02 09:39:57 [debug] 2220074#0: *7 http args: ""
2011/05/02 09:39:57 [debug] 2220074#0: *7 http exten: ""
2011/05/02 09:39:57 [debug] 2220074#0: *7 http process request header
line
2011/05/02 09:39:57 [debug] 2220074#0: *7 http header: "User-Agent:
curl/7.21.6 (amd64-pc-win32) libcurl/7.21.6 OpenSSL/0.9.8r zlib/1.2.5"
2011/05/02 09:39:57 [debug] 2220074#0: *7 http header: "Host:
10.0.4.237"
2011/05/02 09:39:57 [debug] 2220074#0: *7 http header: "Accept: */*"
2011/05/02 09:39:57 [debug] 2220074#0: *7 http header done
2011/05/02 09:39:57 [debug] 2220074#0: *7 event timer del: 3:
2968566242
2011/05/02 09:39:57 [debug] 2220074#0: *7 rewrite phase: 0
2011/05/02 09:39:57 [debug] 2220074#0: *7 test location: "/"
2011/05/02 09:39:57 [debug] 2220074#0: *7 test location: ~ "\.cgi$"
2011/05/02 09:39:57 [debug] 2220074#0: *7 using configuration "/"
2011/05/02 09:39:57 [debug] 2220074#0: *7 http cl:-1 max:1048576
2011/05/02 09:39:57 [debug] 2220074#0: *7 rewrite phase: 2
2011/05/02 09:39:57 [debug] 2220074#0: *7 post rewrite phase: 3
2011/05/02 09:39:57 [debug] 2220074#0: *7 generic phase: 4
2011/05/02 09:39:57 [debug] 2220074#0: *7 generic phase: 5
2011/05/02 09:39:57 [debug] 2220074#0: *7 access phase: 6
2011/05/02 09:39:57 [debug] 2220074#0: *7 access phase: 7
2011/05/02 09:39:57 [debug] 2220074#0: *7 post access phase: 8
2011/05/02 09:39:57 [debug] 2220074#0: *7 content phase: 9
2011/05/02 09:39:57 [debug] 2220074#0: *7 open index
"/www/webdoc/index.html"
2011/05/02 09:39:57 [debug] 2220074#0: *7 internal redirect:
"/index.html?"
2011/05/02 09:39:57 [debug] 2220074#0: *7 rewrite phase: 0
2011/05/02 09:39:57 [debug] 2220074#0: *7 test location: "/"
2011/05/02 09:39:57 [debug] 2220074#0: *7 test location: "50x.html"
2011/05/02 09:39:57 [debug] 2220074#0: *7 test location: ~ "\.cgi$"
2011/05/02 09:39:57 [debug] 2220074#0: *7 using configuration "/"
2011/05/02 09:39:57 [debug] 2220074#0: *7 http cl:-1 max:1048576
2011/05/02 09:39:57 [debug] 2220074#0: *7 rewrite phase: 2
2011/05/02 09:39:57 [debug] 2220074#0: *7 post rewrite phase: 3
2011/05/02 09:39:57 [debug] 2220074#0: *7 generic phase: 4
2011/05/02 09:39:57 [debug] 2220074#0: *7 generic phase: 5
2011/05/02 09:39:57 [debug] 2220074#0: *7 access phase: 6
2011/05/02 09:39:57 [debug] 2220074#0: *7 access phase: 7
2011/05/02 09:39:57 [debug] 2220074#0: *7 post access phase: 8
2011/05/02 09:39:57 [debug] 2220074#0: *7 content phase: 9
2011/05/02 09:39:57 [debug] 2220074#0: *7 content phase: 10
2011/05/02 09:39:57 [debug] 2220074#0: *7 content phase: 11
2011/05/02 09:39:57 [debug] 2220074#0: *7 http filename:
"/www/webdoc/index.html"
2011/05/02 09:39:57 [debug] 2220074#0: *7 add cleanup: 080F9748
2011/05/02 09:39:57 [debug] 2220074#0: *7 http static fd: 4
2011/05/02 09:39:57 [debug] 2220074#0: *7 http set discard body
2011/05/02 09:39:57 [debug] 2220074#0: *7 HTTP/1.1 200 OK
Server: nginx/0.8.53
Date: Mon, 02 May 2011 13:39:57 GMT
Content-Type: text/html
Content-Length: 5222
Last-Modified: Wed, 27 Apr 2011 20:25:09 GMT
Connection: keep-alive
Accept-Ranges: bytes

2011/05/02 09:39:57 [debug] 2220074#0: *7 write new buf t:1 f:0
080F981C, pos 080F981C, size: 217 file: 0, size: 0
2011/05/02 09:39:57 [debug] 2220074#0: *7 http write filter: l:0 f:0
s:217
2011/05/02 09:39:57 [debug] 2220074#0: *7 http output filter
"/index.html?"
2011/05/02 09:39:57 [debug] 2220074#0: *7 http copy filter:
"/index.html?"
2011/05/02 09:39:57 [debug] 2220074#0: *7 malloc: 0814BB80:5222
2011/05/02 09:39:57 [debug] 2220074#0: *7 read: 4, 0814BB80, 5222, 0
2011/05/02 09:39:57 [debug] 2220074#0: *7 http postpone filter
"/index.html?" 080F9974
2011/05/02 09:39:57 [debug] 2220074#0: *7 write old buf t:1 f:0
080F981C, pos 080F981C, size: 217 file: 0, size: 0
2011/05/02 09:39:57 [debug] 2220074#0: *7 write new buf t:1 f:0
0814BB80, pos 0814BB80, size: 5222 file: 0, size: 0
2011/05/02 09:39:57 [debug] 2220074#0: *7 http write filter: l:1 f:0
s:5439
2011/05/02 09:39:57 [debug] 2220074#0: *7 http write filter limit 0
2011/05/02 09:39:57 [debug] 2220074#0: *7 posix_memalign: 080E1370:256
@16
2011/05/02 09:39:57 [debug] 2220074#0: *7 malloc: 080FCFF0:16384
2011/05/02 09:39:57 [debug] 2220074#0: *7 http write filter 080F98FC
2011/05/02 09:39:57 [debug] 2220074#0: *7 http copy filter: -2
"/index.html?"
2011/05/02 09:39:57 [debug] 2220074#0: *7 http finalize request: -2,
"/index.html?" a:1, c:2
2011/05/02 09:39:57 [debug] 2220074#0: *7 event timer add: 3:
60000:2968566732
2011/05/02 09:39:57 [debug] 2220074#0: *7 http finalize request: -4,
"/index.html?" a:1, c:2
2011/05/02 09:39:57 [debug] 2220074#0: *7 http request count:2 blk:0

2011/05/02 09:40:57 [debug] 2220074#0: *7 event timer del: 3:
2968566732
2011/05/02 09:40:57 [debug] 2220074#0: *7 http run request:
"/index.html?"
2011/05/02 09:40:57 [debug] 2220074#0: *7 http writer handler:
"/index.html?"
2011/05/02 09:40:57 [info] 2220074#0: *7 client timed out (260:
Connection timed out) while sending response to client, client:
10.0.4.118, server: tf2000, request: "GET / HTTP/1.1", host:
"10.0.4.237"
2011/05/02 09:40:57 [debug] 2220074#0: *7 http finalize request: 408,
"/index.html?" a:1, c:1
2011/05/02 09:40:57 [debug] 2220074#0: *7 http terminate request count:1

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195329,195329#msg-195329



From lists at ruby-forum.com  Wed May  4 01:00:52 2011
From: lists at ruby-forum.com (Mike P.)
Date: Tue, 03 May 2011 23:00:52 +0200
Subject: 502 errors, upstream sent too big header
In-Reply-To: <BANLkTi=CkxwWv5aW8F+Rxn6DCanaebLnhQ@mail.gmail.com>
References: <BANLkTi=CkxwWv5aW8F+Rxn6DCanaebLnhQ@mail.gmail.com>
Message-ID: <706c3a2663e40773c575d47b54ec69aa@ruby-forum.com>

Maxim,

Thanks! This makes a lot more sense now. I'll pursue a resolution with
passenger at this time. Hopefully this is something that they can
eventually resolve with a future release.

Mike P.

-- 
Posted via http://www.ruby-forum.com/.


From kworthington at gmail.com  Wed May  4 06:36:02 2011
From: kworthington at gmail.com (Kevin Worthington)
Date: Tue, 3 May 2011 22:36:02 -0400
Subject: Nginx-1.0.1 for Windows 32-bit and 64-bit now available
Message-ID: <BANLkTinynxpz89F01-o46aOPAjVQ2=gEoQ@mail.gmail.com>

For those interested, Cygwin based versions of Nginx for Windows (both
32-bit and 64-bit) builds are available here:
http://www.kevinworthington.com/nginx-for-windows/

Blog post/announcement: http://www.kevinworthington.com/nginx-101-windows/

Cygwin is NOT required on your system to run this version; all needed
files are included.

Please note that I have been providing Cygwin based builds since
before official Windows builds were made available. It's mostly to
support folks already using the builds.

Thank you,
Kevin
--
Kevin Worthington
http://www.kevinworthington.com/


From agentzh at gmail.com  Wed May  4 06:54:08 2011
From: agentzh at gmail.com (agentzh)
Date: Wed, 4 May 2011 10:54:08 +0800
Subject: Catch ALL requests by LUA-script
In-Reply-To: <90c728b1bd4f045a01171963ba75b30f.NginxMailingListEnglish@forum.nginx.org>
References: <AANLkTimcqUR5y2Yi2e4B5j5k+yZ6GDOvxRmT-Qm4M2NH@mail.gmail.com>
	<eb9c211f9f74635a1e0afcc9d0591e22.NginxMailingListEnglish@forum.nginx.org>
	<90c728b1bd4f045a01171963ba75b30f.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTinB3WNOQ4=pP6nf1x1zNUwc6vVafw@mail.gmail.com>

On Mon, May 2, 2011 at 12:11 PM, HowToMeetLadies <nginx-forum at nginx.us> wrote:
> Maybe slightly off-topic but then maybe it is useful for you (andiL) or
> someone else.
>

Your feedback is always welcome! Sorry for the delay on my side for I
was enjoying my vocation in the Fuzhou city in the southern region of
China in the last few days :)

> Before i start, @agentzh: every ?nth request, one of nginx
> (0.8.54.4rc2-jit2beta6) worker processes segfaults, is this behaviour
> related to the pcre issue? have not patched it yet.. ah and thank you
> *very* much for giving nginx this all new powers!! ((:
>

What's your operating system? We used to observe similar periodic
segfaults with exactly the same version of ngx_openresty on RedHat
Enterprise Linux (RHEL) 5.4 in production but now all is fine after we
switched LuaJIT to a more recent git HEAD version which includes an
important work-around for an old libgcc bug explained here in the
lua-l mailing list:

    http://lua-users.org/lists/lua-l/2011-03/msg00576.html

I'll make a new release of ngx_openresty to include a new version of
LuaJIT as well as some other modules' updates like ngx_lua's and
ngx_srcache's. I'll inform you here when I'm done.

Also, it'll be very appreciated if you can obtain a gdb backtrace for
your segfaults so that we can look at the details ;)

> Ok, im a long time nginx/redis/ruby user and this are my first steps
> with lua and a lightweight (M)VC framework around three core "classes",
> which uses haml for templates/layout and markdown for text processing (5
> minute ffi discount binding), drizzle and redis for data and redis also
> for caching. Successfully running a little app with it. yay! *.* For now
> i can give you some snippets just to show some basic aspects, ill
> release a fully working example sometime soon, if the segfaults doesnt
> belong to the pcre-issue ill try to debug that too.
>

Oh, that's awesome :)

>
> Below ill use luasql, instead you can use
> ngx.location.capture("/some-internal-rds-json") and lua-yajl, but this
> approach ends as soon as you want INSERT followed by LAST_INSERT_ID() or
> other complex queries/transactions (as long as libdrizzle has no support
> for multiple queries, maybe this is not the best option to go with, ive
> tried alternatives and workarounds, but for now this is what *I* want).
>

The luasql library will certainly block nginx worker processes. Please
dont' use that.

I know libdrizzle is very limited in terms of functionality and it's
*not* maintained very well. We'll start a completely new nginx
upstream modules for mysql named ngx_mysql by implementing a
non-blocking full-fledged highly-optimized mysql client from scratch.
I'd love to see it happen this year ;)

Another direction we're taking is to introduce the "cosocket"
mechanism into ngx_lua such that one can use a *synchronous* Lua
socket API that is transparently *asynchronous* and *non-blocking* on
the nginx C level. Atop that, we can build pure-Lua non-blocking mysql
client driver and thus putting everything onto the Lua land and no
longer need to mess up the nginx.conf file and nginx subrequests. I'd
also want to see it happen by this Christmas ;)

> At the moment i do not need an ORM for my data, but have already done
> some functions for both rds_json and luasql. redis has some commands
> exposed in nginx.conf for easy access with lua (: The core classes:
>
> Request; which contains several ngx.var wrappers (for nil instead of ''
> for empty keys), present(str), simple cookie/session/login, before/after
> filters, haml/markdown rendering, etc..
>
> MethodRequest < Request; which calls functions for HTTP Methods aka GET,
> POST alike..
> and finally Resource < MethodRequest; which has an
> Rails-ActionController-like interface.
>

The high-level abstraction for ngx_lua is excellent! We're taking a
slightly different approach along the line in our web applications:

    http://agentzh.org/misc/nginx/lzsql-manual.html

We're using a DSL (Domain Specific Language) compiler that emits Lua
code and business logic is mostly implemented in this DSL. It's
targeting our business though, not meant to be useful for general web
application development (yet).

Thanks for sharing!

-agentzh


From nginx-forum at nginx.us  Wed May  4 07:10:53 2011
From: nginx-forum at nginx.us (stalkercn)
Date: Tue, 03 May 2011 23:10:53 -0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <ccbf8793be4292740ff1b275a855d614.NginxMailingListEnglish@forum.nginx.org>

anyone can help me?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,192785,195465#msg-195465



From lists at ruby-forum.com  Wed May  4 09:24:37 2011
From: lists at ruby-forum.com (Yanxin Z.)
Date: Wed, 04 May 2011 07:24:37 +0200
Subject: limit connection
Message-ID: <3eb40128cef9709ed44491c6d10d7e09@ruby-forum.com>

Hello,
I want to limit the number of connection per hour from IP range.

For example, I want to limit every hour, each IP can only visit our
website for 5 times.

How can I write the config?

There is limit_req, which limit the connection rate, used for DDos
control.
But it is different from my requirement.

Thanks,
Yanxin

-- 
Posted via http://www.ruby-forum.com/.


From caldcv at gmail.com  Wed May  4 09:27:57 2011
From: caldcv at gmail.com (Chris)
Date: Wed, 4 May 2011 01:27:57 -0400
Subject: limit connection
In-Reply-To: <3eb40128cef9709ed44491c6d10d7e09@ruby-forum.com>
References: <3eb40128cef9709ed44491c6d10d7e09@ruby-forum.com>
Message-ID: <BANLkTikiFAtQHofO98QpF-14OodNOvgi=Q@mail.gmail.com>

You might be better off researching a solution with iptables

On 5/4/11, Yanxin Z. <lists at ruby-forum.com> wrote:
> Hello,
> I want to limit the number of connection per hour from IP range.
>
> For example, I want to limit every hour, each IP can only visit our
> website for 5 times.
>
> How can I write the config?
>
> There is limit_req, which limit the connection rate, used for DDos
> control.
> But it is different from my requirement.
>
> Thanks,
> Yanxin
>
> --
> Posted via http://www.ruby-forum.com/.
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>


-- 
--C

"The dumber people think you are, the more surprised they're going to
be when you kill them." - Sir William Clayton


From igor at sysoev.ru  Wed May  4 10:26:50 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 4 May 2011 10:26:50 +0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <ccbf8793be4292740ff1b275a855d614.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
	<ccbf8793be4292740ff1b275a855d614.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110504062650.GA57647@sysoev.ru>

On Tue, May 03, 2011 at 11:10:53PM -0400, stalkercn wrote:
> anyone can help me?

Could you show "nginx -V" output ?


-- 
Igor Sysoev


From igor at sysoev.ru  Wed May  4 15:08:17 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 4 May 2011 15:08:17 +0400
Subject: POSIX semaphores patch
Message-ID: <20110504110817.GE72381@sysoev.ru>

The attache patch enables POSIX semaphores usage during shared memory access.
It should decrease CPU usage on start-up or online upgrade of installations
with large caches or large number of workers.

I tested the patch on FreeBSD 8.2. There is no sense to test it on
FreeBSD prior to 7.2 version, since semaphores does not work on these systems.
You can ensure that there are enough semaphores:
   sysctl p1003_1b.sem_nsems_max
By default there should be 32 semaphores. On 8.2+ you can increase them
dynamically. You need one semaphores per shared zone.
"sysctl p1003_1b.nsems" shows current number of semaphores.

It would be intresting to test the patch on other platforms - Linux
and Solaris. Please draw attention to alerts in error_log.
Note also that nginx does not destroy semaphores on exit. They should
deleted by kernel automatically. Please check that there is no
semaphore leakage while nginx work.


-- 
Igor Sysoev
-------------- next part --------------
Index: src/event/ngx_event.c
===================================================================
--- src/event/ngx_event.c	(revision 3907)
+++ src/event/ngx_event.c	(working copy)
@@ -519,6 +519,7 @@
     shared = shm.addr;
 
     ngx_accept_mutex_ptr = (ngx_atomic_t *) shared;
+    ngx_accept_mutex.spin = (ngx_uint_t) -1;
 
     if (ngx_shmtx_create(&ngx_accept_mutex, shared, cycle->lock_file.data)
         != NGX_OK)
Index: src/os/unix/ngx_posix_config.h
===================================================================
--- src/os/unix/ngx_posix_config.h	(revision 3907)
+++ src/os/unix/ngx_posix_config.h	(working copy)
@@ -96,6 +96,11 @@
 #include <ngx_auto_config.h>
 
 
+#if (NGX_HAVE_POSIX_SEM)
+#include <semaphore.h>
+#endif
+
+
 #if (NGX_HAVE_POLL)
 #include <poll.h>
 #endif
Index: src/os/unix/ngx_linux_config.h
===================================================================
--- src/os/unix/ngx_linux_config.h	(revision 3907)
+++ src/os/unix/ngx_linux_config.h	(working copy)
@@ -58,6 +58,11 @@
 #include <ngx_auto_config.h>
 
 
+#if (NGX_HAVE_POSIX_SEM)
+#include <semaphore.h>
+#endif
+
+
 #if (NGX_HAVE_SYS_PRCTL_H)
 #include <sys/prctl.h>
 #endif
Index: src/os/unix/ngx_freebsd_config.h
===================================================================
--- src/os/unix/ngx_freebsd_config.h	(revision 3907)
+++ src/os/unix/ngx_freebsd_config.h	(working copy)
@@ -68,6 +68,11 @@
 #include <ngx_auto_config.h>
 
 
+#if (NGX_HAVE_POSIX_SEM)
+#include <semaphore.h>
+#endif
+
+
 #if (NGX_HAVE_POLL)
 #include <poll.h>
 #endif
Index: src/os/unix/ngx_solaris_config.h
===================================================================
--- src/os/unix/ngx_solaris_config.h	(revision 3907)
+++ src/os/unix/ngx_solaris_config.h	(working copy)
@@ -57,6 +57,11 @@
 #include <ngx_auto_config.h>
 
 
+#if (NGX_HAVE_POSIX_SEM)
+#include <semaphore.h>
+#endif
+
+
 #if (NGX_HAVE_POLL)
 #include <poll.h>
 #endif
Index: src/os/unix/ngx_darwin_config.h
===================================================================
--- src/os/unix/ngx_darwin_config.h	(revision 3907)
+++ src/os/unix/ngx_darwin_config.h	(working copy)
@@ -56,6 +56,11 @@
 #include <ngx_auto_config.h>
 
 
+#if (NGX_HAVE_POSIX_SEM)
+#include <semaphore.h>
+#endif
+
+
 #if (NGX_HAVE_POLL)
 #include <poll.h>
 #endif
Index: src/core/ngx_shmtx.c
===================================================================
--- src/core/ngx_shmtx.c	(revision 3907)
+++ src/core/ngx_shmtx.c	(working copy)
@@ -16,9 +16,160 @@
 {
     mtx->lock = addr;
 
+    if (mtx->spin == (ngx_uint_t) -1) {
+        return NGX_OK;
+    }
+
+    mtx->spin = 2048;
+
+#if (NGX_HAVE_POSIX_SEM)
+
+    if (sem_init(&mtx->sem, 1, 0) == -1) {
+        ngx_log_error(NGX_LOG_ALERT, ngx_cycle->log, ngx_errno,
+                      "sem_init() failed");
+    } else {
+        mtx->semaphore = 1;
+    }
+
+#endif
+
     return NGX_OK;
 }
 
+
+void
+ngx_shmtx_destory(ngx_shmtx_t *mtx)
+{
+#if (NGX_HAVE_POSIX_SEM)
+
+    if (mtx->semaphore) {
+        if (sem_destroy(&mtx->sem) == -1) {
+            ngx_log_error(NGX_LOG_ALERT, ngx_cycle->log, ngx_errno,
+                          "sem_destroy() failed");
+        }
+    }
+
+#endif
+}
+
+
+ngx_uint_t
+ngx_shmtx_trylock(ngx_shmtx_t *mtx)
+{
+    ngx_atomic_uint_t  val;
+
+    val = *mtx->lock;
+
+    return ((val & 0x80000000) == 0
+            && ngx_atomic_cmp_set(mtx->lock, val, val | 0x80000000));
+}
+
+
+void
+ngx_shmtx_lock(ngx_shmtx_t *mtx)
+{
+    ngx_uint_t         i, n;
+    ngx_atomic_uint_t  val;
+
+    ngx_log_debug0(NGX_LOG_DEBUG_CORE, ngx_cycle->log, 0, "shmtx lock");
+
+    for ( ;; ) {
+
+        val = *mtx->lock;
+
+        if ((val & 0x80000000) == 0
+            && ngx_atomic_cmp_set(mtx->lock, val, val | 0x80000000))
+        {
+            return;
+        }
+
+        if (ngx_ncpu > 1) {
+
+            for (n = 1; n < mtx->spin; n <<= 1) {
+
+                for (i = 0; i < n; i++) {
+                    ngx_cpu_pause();
+                }
+
+                val = *mtx->lock;
+
+                if ((val & 0x80000000) == 0
+                    && ngx_atomic_cmp_set(mtx->lock, val, val | 0x80000000))
+                {
+                    return;
+                }
+            }
+        }
+
+#if (NGX_HAVE_POSIX_SEM)
+
+        if (mtx->semaphore) {
+            val = *mtx->lock;
+
+            if ((val & 0x80000000)
+                && ngx_atomic_cmp_set(mtx->lock, val, val + 1))
+            {
+                ngx_log_debug1(NGX_LOG_DEBUG_CORE, ngx_cycle->log, 0,
+                               "shmtx wait %XA", val);
+
+                if (sem_wait(&mtx->sem) == -1) {
+                    ngx_log_error(NGX_LOG_ALERT, ngx_cycle->log, ngx_errno,
+                                  "sem_wait() failed while waiting on shmtx");
+                }
+
+                ngx_log_debug0(NGX_LOG_DEBUG_CORE, ngx_cycle->log, 0,
+                               "shmtx awoke");
+
+                continue;
+            }
+        }
+
+#endif
+
+        ngx_sched_yield();
+    }
+}
+
+
+void
+ngx_shmtx_unlock(ngx_shmtx_t *mtx)
+{
+    ngx_atomic_uint_t  val, old, wait;
+
+    if (mtx->spin != (ngx_uint_t) -1) {
+        ngx_log_debug0(NGX_LOG_DEBUG_CORE, ngx_cycle->log, 0, "shmtx unlock");
+    }
+
+    for ( ;; ) {
+
+        old = *mtx->lock;
+        wait = old & 0x7fffffff;
+        val = wait ? wait - 1 : 0;
+
+        if (ngx_atomic_cmp_set(mtx->lock, old, val)) {
+            break;
+        }
+    }
+
+#if (NGX_HAVE_POSIX_SEM)
+    {
+
+    if (wait == 0 || !mtx->semaphore) {
+        return;
+    }
+
+    ngx_log_debug1(NGX_LOG_DEBUG_CORE, ngx_cycle->log, 0,
+                   "shmtx wake %XA", old);
+
+    if (sem_post(&mtx->sem) == -1) {
+        ngx_log_error(NGX_LOG_ALERT, ngx_cycle->log, ngx_errno,
+                      "sem_post() failed while wake shmtx");
+    }
+    }
+#endif
+}
+
+
 #else
 
 
@@ -65,4 +216,62 @@
 }
 
 
+ngx_uint_t
+ngx_shmtx_trylock(ngx_shmtx_t *mtx)
+{
+    ngx_err_t  err;
+
+    err = ngx_trylock_fd(mtx->fd);
+
+    if (err == 0) {
+        return 1;
+    }
+
+    if (err == NGX_EAGAIN) {
+        return 0;
+    }
+
+#if __osf__ /* Tru64 UNIX */
+
+    if (err == NGX_EACCESS) {
+        return 0;
+    }
+
 #endif
+
+    ngx_log_abort(err, ngx_trylock_fd_n " %s failed", mtx->name);
+
+    return 0;
+}
+
+
+void
+ngx_shmtx_lock(ngx_shmtx_t *mtx)
+{
+    ngx_err_t  err;
+
+    err = ngx_lock_fd(mtx->fd);
+
+    if (err == 0) {
+        return;
+    }
+
+    ngx_log_abort(err, ngx_lock_fd_n " %s failed", mtx->name);
+}
+
+
+void
+ngx_shmtx_unlock(ngx_shmtx_t *mtx)
+{
+    ngx_err_t  err;
+
+    err = ngx_unlock_fd(mtx->fd);
+
+    if (err == 0) {
+        return;
+    }
+
+    ngx_log_abort(err, ngx_unlock_fd_n " %s failed", mtx->name);
+}
+
+#endif
Index: src/core/ngx_shmtx.h
===================================================================
--- src/core/ngx_shmtx.h	(revision 3907)
+++ src/core/ngx_shmtx.h	(working copy)
@@ -15,95 +15,23 @@
 typedef struct {
 #if (NGX_HAVE_ATOMIC_OPS)
     ngx_atomic_t  *lock;
+#if (NGX_HAVE_POSIX_SEM)
+    ngx_uint_t     semaphore;
+    sem_t          sem;
+#endif
 #else
     ngx_fd_t       fd;
     u_char        *name;
 #endif
+    ngx_uint_t     spin;
 } ngx_shmtx_t;
 
 
 ngx_int_t ngx_shmtx_create(ngx_shmtx_t *mtx, void *addr, u_char *name);
-
-
-#if (NGX_HAVE_ATOMIC_OPS)
-
-static ngx_inline ngx_uint_t
-ngx_shmtx_trylock(ngx_shmtx_t *mtx)
-{
-    return (*mtx->lock == 0 && ngx_atomic_cmp_set(mtx->lock, 0, ngx_pid));
-}
-
-#define ngx_shmtx_lock(mtx)   ngx_spinlock((mtx)->lock, ngx_pid, 1024)
-
-#define ngx_shmtx_unlock(mtx) (void) ngx_atomic_cmp_set((mtx)->lock, ngx_pid, 0)
-
-#define ngx_shmtx_destory(mtx)
-
-
-#else
-
-static ngx_inline ngx_uint_t
-ngx_shmtx_trylock(ngx_shmtx_t *mtx)
-{
-    ngx_err_t  err;
-
-    err = ngx_trylock_fd(mtx->fd);
-
-    if (err == 0) {
-        return 1;
-    }
-
-    if (err == NGX_EAGAIN) {
-        return 0;
-    }
-
-#if __osf__ /* Tru64 UNIX */
-
-    if (err == NGX_EACCESS) {
-        return 0;
-    }
-
-#endif
-
-    ngx_log_abort(err, ngx_trylock_fd_n " %s failed", mtx->name);
-
-    return 0;
-}
-
-
-static ngx_inline void
-ngx_shmtx_lock(ngx_shmtx_t *mtx)
-{
-    ngx_err_t  err;
-
-    err = ngx_lock_fd(mtx->fd);
-
-    if (err == 0) {
-        return;
-    }
-
-    ngx_log_abort(err, ngx_lock_fd_n " %s failed", mtx->name);
-}
-
-
-static ngx_inline void
-ngx_shmtx_unlock(ngx_shmtx_t *mtx)
-{
-    ngx_err_t  err;
-
-    err = ngx_unlock_fd(mtx->fd);
-
-    if (err == 0) {
-        return;
-    }
-
-    ngx_log_abort(err, ngx_unlock_fd_n " %s failed", mtx->name);
-}
-
-
 void ngx_shmtx_destory(ngx_shmtx_t *mtx);
+ngx_uint_t ngx_shmtx_trylock(ngx_shmtx_t *mtx);
+void ngx_shmtx_lock(ngx_shmtx_t *mtx);
+void ngx_shmtx_unlock(ngx_shmtx_t *mtx);
 
-#endif
 
-
 #endif /* _NGX_SHMTX_H_INCLUDED_ */
Index: auto/os/freebsd
===================================================================
--- auto/os/freebsd	(revision 3907)
+++ auto/os/freebsd	(working copy)
@@ -49,7 +49,16 @@
     have=NGX_HAVE_AIO_SENDFILE . auto/have
 fi
 
+# POSIX semaphores
+# http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/127545
 
+if [ $osreldate -ge 701106 ]; then
+    echo " + POSIX semaphores should work"
+else
+    have=NGX_HAVE_POSIX_SEM . auto/nohave
+fi
+
+
 # kqueue
 
 if [ \( $osreldate -lt 500000 -a $osreldate -ge 410000 \) \
Index: auto/unix
===================================================================
--- auto/unix	(revision 3907)
+++ auto/unix	(working copy)
@@ -234,6 +234,18 @@
 . auto/feature
 
 
+ngx_feature="POSIX semaphores"
+ngx_feature_name="NGX_HAVE_POSIX_SEM"
+ngx_feature_run=yes
+ngx_feature_incs="#include <semaphore.h>"
+ngx_feature_path=
+ngx_feature_libs=
+ngx_feature_test="sem_t  sem;
+                  if (sem_init(&sem, 1, 0) == -1) return 1;
+                  sem_destroy(&sem);"
+. auto/feature
+
+
 ngx_feature="struct msghdr.msg_control"
 ngx_feature_name="NGX_HAVE_MSGHDR_MSG_CONTROL"
 ngx_feature_run=no

From nginx-forum at nginx.us  Wed May  4 15:09:21 2011
From: nginx-forum at nginx.us (burn)
Date: Wed, 04 May 2011 07:09:21 -0400
Subject: not working without aio
Message-ID: <a36328c7a87ec15e9ce4b96575f7436a.NginxMailingListEnglish@forum.nginx.org>

If you install nginx compiled with aio support on non-aio enabled
kernel, it doesn't work:

2011/04/15 11:46:25 [emerg] 18942#0: io_setup() failed (1: Operation not
permitted)
2011/04/15 11:46:25 [alert] 18938#0: worker process 18942 exited with
fatal code 2 and can not be respawn

All workers die, only master process remains, requests aren't served

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195522,195522#msg-195522



From kgorlo at gmail.com  Wed May  4 15:25:38 2011
From: kgorlo at gmail.com (Kamil Gorlo)
Date: Wed, 4 May 2011 13:25:38 +0200
Subject: Load balancing based on part of url or header
Message-ID: <BANLkTim121GKJD-AvUine3G-cbZLkpacJA@mail.gmail.com>

Hi guys,

is this possible to have similar feature as ip_hash but not based on
IP address but on some part of request url (or header)?

Cheers,
--
Kamil Gorlo


From mdounin at mdounin.ru  Wed May  4 15:27:33 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Wed, 4 May 2011 15:27:33 +0400
Subject: not working without aio
In-Reply-To: <a36328c7a87ec15e9ce4b96575f7436a.NginxMailingListEnglish@forum.nginx.org>
References: <a36328c7a87ec15e9ce4b96575f7436a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110504112733.GT42265@mdounin.ru>

Hello!

On Wed, May 04, 2011 at 07:09:21AM -0400, burn wrote:

> If you install nginx compiled with aio support on non-aio enabled
> kernel, it doesn't work:
> 
> 2011/04/15 11:46:25 [emerg] 18942#0: io_setup() failed (1: Operation not
> permitted)
> 2011/04/15 11:46:25 [alert] 18938#0: worker process 18942 exited with
> fatal code 2 and can not be respawn
> 
> All workers die, only master process remains, requests aren't served

You have to recompile nginx without aio.  Run-time fallback isn't 
currently supported.

Maxim Dounin


From mdounin at mdounin.ru  Wed May  4 15:28:29 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Wed, 4 May 2011 15:28:29 +0400
Subject: Load balancing based on part of url or header
In-Reply-To: <BANLkTim121GKJD-AvUine3G-cbZLkpacJA@mail.gmail.com>
References: <BANLkTim121GKJD-AvUine3G-cbZLkpacJA@mail.gmail.com>
Message-ID: <20110504112828.GU42265@mdounin.ru>

Hello!

On Wed, May 04, 2011 at 01:25:38PM +0200, Kamil Gorlo wrote:

> Hi guys,
> 
> is this possible to have similar feature as ip_hash but not based on
> IP address but on some part of request url (or header)?

There are 3rd party modules available which do this, check wiki.

Maxim Dounin


From lists at ruby-forum.com  Wed May  4 19:38:56 2011
From: lists at ruby-forum.com (Yanxin Z.)
Date: Wed, 04 May 2011 17:38:56 +0200
Subject: limit connection
In-Reply-To: <3eb40128cef9709ed44491c6d10d7e09@ruby-forum.com>
References: <3eb40128cef9709ed44491c6d10d7e09@ruby-forum.com>
Message-ID: <d05379bf98bb9884130758562db2ea69@ruby-forum.com>

Thanks Chris.
Do you have example I can refer?
Yanxin

-- 
Posted via http://www.ruby-forum.com/.


From caldcv at gmail.com  Wed May  4 20:36:39 2011
From: caldcv at gmail.com (Chris)
Date: Wed, 4 May 2011 12:36:39 -0400
Subject: limit connection
In-Reply-To: <d05379bf98bb9884130758562db2ea69@ruby-forum.com>
References: <3eb40128cef9709ed44491c6d10d7e09@ruby-forum.com>
	<d05379bf98bb9884130758562db2ea69@ruby-forum.com>
Message-ID: <BANLkTincrAdzqF8QRNqgFbb3fXtd6Sf0nA@mail.gmail.com>

Many are available online in regards to SSH limiting however you have
to put in IP ranges of people you do not want to limit, which could be
very hard. What exactly are you trying to do?

On 5/4/11, Yanxin Z. <lists at ruby-forum.com> wrote:
> Thanks Chris.
> Do you have example I can refer?
> Yanxin
>
> --
> Posted via http://www.ruby-forum.com/.
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>


-- 
--C

"The dumber people think you are, the more surprised they're going to
be when you kill them." - Sir William Clayton


From clanherb at gmail.com  Wed May  4 22:59:21 2011
From: clanherb at gmail.com (=?UTF-8?B?5p2o6ZWt?=)
Date: Thu, 5 May 2011 02:59:21 +0800
Subject: question about the context of map directive
Message-ID: <BANLkTinkf+Dsj3S7bCGd-LwPXnwwne7H6g@mail.gmail.com>

Hi,

The context of map directive is *http* as documented on wiki.nginx.org. This
means all the requests will go through the map filter phase.
This is not ideal when you have many virtual hosts.

For example, we've 50+ virtual hosts. When one of them need map variables,
the others automatically inherited those settings because there is no way to
limit it to server context.

If this is the situation, can someone know the internals of nginx explain
why the map module can't work at server level.
Thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110505/b46aa5fb/attachment.html>

From Brian.Akins at turner.com  Wed May  4 23:03:33 2011
From: Brian.Akins at turner.com (Akins, Brian)
Date: Wed, 04 May 2011 15:03:33 -0400
Subject: Lua pcre inside nginx
Message-ID: <C9E71B45.DD104%Brian.Akins@turner.com>

I wanted to use pcre inside a lua script using rex_pcre.  However, I always
get the error:

failed to get memory (pattern offset: 15)

The snippet of code always works in a standalone script.

Ideas?

-- 
Brian Akins



From igor at sysoev.ru  Wed May  4 23:10:00 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 4 May 2011 23:10:00 +0400
Subject: question about the context of map directive
In-Reply-To: <BANLkTinkf+Dsj3S7bCGd-LwPXnwwne7H6g@mail.gmail.com>
References: <BANLkTinkf+Dsj3S7bCGd-LwPXnwwne7H6g@mail.gmail.com>
Message-ID: <20110504191000.GA96440@sysoev.ru>

On Thu, May 05, 2011 at 02:59:21AM +0800, ?? wrote:
> Hi,
> 
> The context of map directive is *http* as documented on wiki.nginx.org. This
> means all the requests will go through the map filter phase.
> This is not ideal when you have many virtual hosts.
> 
> For example, we've 50+ virtual hosts. When one of them need map variables,
> the others automatically inherited those settings because there is no way to
> limit it to server context.
> 
> If this is the situation, can someone know the internals of nginx explain
> why the map module can't work at server level.
> Thanks.

The most nginx directives are declarative. This means that nginx
looks up a map only when it needs a variable value defined via the map.
There are directives that are really executed: "rewrite" and "if".
Avoid them.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Wed May  4 23:10:38 2011
From: nginx-forum at nginx.us (art.wu)
Date: Wed, 04 May 2011 15:10:38 -0400
Subject: no server response in https
In-Reply-To: <f2292582eeb12229fb3a95a01b37b61e.NginxMailingListEnglish@forum.nginx.org>
References: <f2292582eeb12229fb3a95a01b37b61e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <0338c925ce50fa5f3695412caf42853a.NginxMailingListEnglish@forum.nginx.org>

There was no response from the server (no 408 received by client).
Thanks for your help.

config:


user   nginx  nginx;
worker_processes  1;

events {
    worker_connections  256;
}


http {
    include       /usr/pkg/etc/nginx/mime.types;
    default_type  application/octet-stream;

    access_log /var/log/nginx/access.log;
    error_log /var/log/nginx/error.log debug;

    sendfile        on;

    keepalive_timeout  65;


    # HTTPS server
    #

    server {
        listen       443;
        server_name  tf2000;

        ssl                             on;
	ssl_certificate          /x509/server.crt;
        ssl_certificate_key  /x509/server.key;

        ssl_session_timeout  5m;

        ssl_protocols  SSLv2 SSLv3 TLSv1;
        ssl_ciphers 
ALL:!ADH:!EXPORT56:RC4+RSA:+HIGH:+MEDIUM:+LOW:+SSLv2:+EXP;
        ssl_prefer_server_ciphers   on;

        location / {
            root   /www/webdoc;
            index  index.html index.htm;
        }
    }
}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195329,195629#msg-195629



From igor at sysoev.ru  Wed May  4 23:15:22 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 4 May 2011 23:15:22 +0400
Subject: question about the context of map directive
In-Reply-To: <20110504191000.GA96440@sysoev.ru>
References: <BANLkTinkf+Dsj3S7bCGd-LwPXnwwne7H6g@mail.gmail.com>
	<20110504191000.GA96440@sysoev.ru>
Message-ID: <20110504191522.GC96440@sysoev.ru>

On Wed, May 04, 2011 at 11:10:00PM +0400, Igor Sysoev wrote:
> On Thu, May 05, 2011 at 02:59:21AM +0800, ?? wrote:
> > Hi,
> > 
> > The context of map directive is *http* as documented on wiki.nginx.org. This
> > means all the requests will go through the map filter phase.
> > This is not ideal when you have many virtual hosts.
> > 
> > For example, we've 50+ virtual hosts. When one of them need map variables,
> > the others automatically inherited those settings because there is no way to
> > limit it to server context.
> > 
> > If this is the situation, can someone know the internals of nginx explain
> > why the map module can't work at server level.
> > Thanks.
> 
> The most nginx directives are declarative. This means that nginx
> looks up a map only when it needs a variable value defined via the map.
> There are directives that are really executed: "rewrite" and "if".
> Avoid them.

Also "set" and "break".


-- 
Igor Sysoev


From mdounin at mdounin.ru  Thu May  5 00:35:47 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 5 May 2011 00:35:47 +0400
Subject: question about the context of map directive
In-Reply-To: <BANLkTinkf+Dsj3S7bCGd-LwPXnwwne7H6g@mail.gmail.com>
References: <BANLkTinkf+Dsj3S7bCGd-LwPXnwwne7H6g@mail.gmail.com>
Message-ID: <20110504203547.GX42265@mdounin.ru>

Hello!

On Thu, May 05, 2011 at 02:59:21AM +0800, ?? wrote:

> Hi,
> 
> The context of map directive is *http* as documented on wiki.nginx.org. This

Yes.

> means all the requests will go through the map filter phase.

No.  Map directive creates a variable and only does something when 
(and if) the variable is accessed.  There is no performance 
penalty for processing requests which doesn't use the variable.

> This is not ideal when you have many virtual hosts.
> 
> For example, we've 50+ virtual hosts. When one of them need map variables,
> the others automatically inherited those settings because there is no way to
> limit it to server context.
> 
> If this is the situation, can someone know the internals of nginx explain
> why the map module can't work at server level.

Changing map module to provide data based on per-server configs is 
possible, but obviously will require extra code.

Maxim Dounin


From pasik at iki.fi  Thu May  5 01:15:50 2011
From: pasik at iki.fi (Pasi =?iso-8859-1?Q?K=E4rkk=E4inen?=)
Date: Thu, 5 May 2011 00:15:50 +0300
Subject: nginx proxying do not locally save/buffer uploaded files?
Message-ID: <20110504211549.GL32595@reaktio.net>

Hello,

I'm running nginx as a proxy in front of some backend servers,
and while uploading files (HTTP PUT) nginx seems to buffer the files
to local disk.

This is not a possible method for this setup, 
because some of the files are so big that they 
won't fit to the local disk of the nginx box.

Is there an option to disable this behaviour?
I'd like to directly stream the uploads to the backend..

Thanks,

-- Pasi



From nginx-forum at nginx.us  Thu May  5 01:35:54 2011
From: nginx-forum at nginx.us (jjjx128)
Date: Wed, 04 May 2011 17:35:54 -0400
Subject: nginx proxying do not locally save/buffer uploaded files?
In-Reply-To: <20110504211549.GL32595@reaktio.net>
References: <20110504211549.GL32595@reaktio.net>
Message-ID: <153c3efe1bd09d636dcea83de2eb5706.NginxMailingListEnglish@forum.nginx.org>

Pasi K?rkk?inen Wrote:
-------------------------------------------------------
> Is there an option to disable this behaviour?
> I'd like to directly stream the uploads to the
> backend..

No. Nginx has no way of streaming data directly to the backend. Take a
look at haproxy or any other tcp load balancer.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195656,195659#msg-195659



From ngx.eugaia at gmail.com  Thu May  5 01:37:11 2011
From: ngx.eugaia at gmail.com (Eugaia)
Date: Thu, 05 May 2011 00:37:11 +0300
Subject: Lua pcre inside nginx
In-Reply-To: <C9E71B45.DD104%Brian.Akins@turner.com>
References: <C9E71B45.DD104%Brian.Akins@turner.com>
Message-ID: <4DC1C707.1000705@gmail.com>

Hi,

On 04/05/2011 22:03, Akins, Brian wrote:
> I wanted to use pcre inside a lua script using rex_pcre.  However, I always
> get the error:
>
> failed to get memory (pattern offset: 15)
>
> The snippet of code always works in a standalone script.
>
> Ideas?
I just tried the supplied test suite for rex_pcre on the latest (as of a 
few days ago) trunk version of LuaJIT 2.0 on Nginx 1.0.0 and there were 
no test failures.  Maybe try JIT if you're not using it already?

Cheers,

Marcus.


From Brian.Akins at turner.com  Thu May  5 01:39:52 2011
From: Brian.Akins at turner.com (Akins, Brian)
Date: Wed, 04 May 2011 17:39:52 -0400
Subject: Lua pcre inside nginx
In-Reply-To: <4DC1C707.1000705@gmail.com>
Message-ID: <C9E73FE8.DD11C%Brian.Akins@turner.com>

On 5/4/11 5:37 PM, "Eugaia" <ngx.eugaia at gmail.com> wrote:


> I just tried the supplied test suite for rex_pcre on the latest (as of a
> few days ago) trunk version of LuaJIT 2.0 on Nginx 1.0.0 and there were
> no test failures.  Maybe try JIT if you're not using it already?
>
I already am using luajit.  I was using rex_pcre from luarocks.


-- 
Brian Akins



From pasik at iki.fi  Thu May  5 01:40:05 2011
From: pasik at iki.fi (Pasi =?iso-8859-1?Q?K=E4rkk=E4inen?=)
Date: Thu, 5 May 2011 00:40:05 +0300
Subject: nginx proxying do not locally save/buffer uploaded files?
In-Reply-To: <153c3efe1bd09d636dcea83de2eb5706.NginxMailingListEnglish@forum.nginx.org>
References: <20110504211549.GL32595@reaktio.net>
	<153c3efe1bd09d636dcea83de2eb5706.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110504214005.GM32595@reaktio.net>

On Wed, May 04, 2011 at 05:35:54PM -0400, jjjx128 wrote:
> Pasi K?rkk?inen Wrote:
> -------------------------------------------------------
> > Is there an option to disable this behaviour?
> > I'd like to directly stream the uploads to the
> > backend..
> 
> No. Nginx has no way of streaming data directly to the backend. Take a
> look at haproxy or any other tcp load balancer.
> 

in tcp mode you can't filter the requests etc.. I need to do that aswell.

Maybe I need to take a look at the code and try to figure out
if direct upload streaming could be implemented in nginx..

Thanks,

-- Pasi



From nginx-forum at nginx.us  Thu May  5 01:56:43 2011
From: nginx-forum at nginx.us (jjjx128)
Date: Wed, 04 May 2011 17:56:43 -0400
Subject: nginx proxying do not locally save/buffer uploaded files?
In-Reply-To: <20110504214005.GM32595@reaktio.net>
References: <20110504214005.GM32595@reaktio.net>
Message-ID: <3dfd6d89a78e3ee124ed58524312dada.NginxMailingListEnglish@forum.nginx.org>

Pasi K?rkk?inen Wrote:
-------------------------------------------------------
> in tcp mode you can't filter the requests etc.. I
> need to do that aswell.

There are quite a few load balancers and proxies out there. I'm sure you
can find something useful.
 
> Maybe I need to take a look at the code and try to
> figure out
> if direct upload streaming could be implemented in
> nginx..

It is possible. So, good luck I guess.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195656,195665#msg-195665



From Brian.Akins at turner.com  Thu May  5 02:11:07 2011
From: Brian.Akins at turner.com (Akins, Brian)
Date: Wed, 04 May 2011 18:11:07 -0400
Subject: Lua pcre inside nginx
In-Reply-To: <C9E73FE8.DD11C%Brian.Akins@turner.com>
Message-ID: <C9E7473B.DD125%Brian.Akins@turner.com>

On 5/4/11 5:39 PM, "Brian Akins" <Brian.Akins at turner.com> wrote:
 
> I already am using luajit.  I was using rex_pcre from luarocks.

I grabbed latest rex_pcre, same issue.  This is x86_64 if that matters.

-- 
Brian Akins



From lists at ruby-forum.com  Thu May  5 02:21:08 2011
From: lists at ruby-forum.com (Yanxin Z.)
Date: Thu, 05 May 2011 00:21:08 +0200
Subject: limit connection
In-Reply-To: <3eb40128cef9709ed44491c6d10d7e09@ruby-forum.com>
References: <3eb40128cef9709ed44491c6d10d7e09@ruby-forum.com>
Message-ID: <bfc40b8d5b85373a048ff5de53dd6774@ruby-forum.com>

Hi Chris,
I want to control the public to do query inside my website, so that bad 
guy can not do database harvest.
For example, each IP can only do query 10 times per day.

The IP I do not want to limit is my own IP. So it's simple to recognize 
the IP range.

Please let me know your opinion.
Thanks,
Yanxin

-- 
Posted via http://www.ruby-forum.com/.


From caldcv at gmail.com  Thu May  5 02:25:19 2011
From: caldcv at gmail.com (Chris)
Date: Wed, 4 May 2011 18:25:19 -0400
Subject: limit connection
In-Reply-To: <bfc40b8d5b85373a048ff5de53dd6774@ruby-forum.com>
References: <3eb40128cef9709ed44491c6d10d7e09@ruby-forum.com>
	<bfc40b8d5b85373a048ff5de53dd6774@ruby-forum.com>
Message-ID: <BANLkTinFDy9NQs4okKRS2Yd=DD8ezvAXBQ@mail.gmail.com>

http://www.shanestillwell.com/blog/2009/03/24/throttle-traffic-iptables

I wish you luck on a solution but it seems rather odd

On 5/4/11, Yanxin Z. <lists at ruby-forum.com> wrote:
> Hi Chris,
> I want to control the public to do query inside my website, so that bad
> guy can not do database harvest.
> For example, each IP can only do query 10 times per day.
>
> The IP I do not want to limit is my own IP. So it's simple to recognize
> the IP range.
>
> Please let me know your opinion.
> Thanks,
> Yanxin
>
> --
> Posted via http://www.ruby-forum.com/.
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>


-- 
--C

"The dumber people think you are, the more surprised they're going to
be when you kill them." - Sir William Clayton


From nginx-forum at nginx.us  Thu May  5 04:05:17 2011
From: nginx-forum at nginx.us (HowToMeetLadies)
Date: Wed, 04 May 2011 20:05:17 -0400
Subject: Catch ALL requests by LUA-script
In-Reply-To: <BANLkTinB3WNOQ4=pP6nf1x1zNUwc6vVafw@mail.gmail.com>
References: <BANLkTinB3WNOQ4=pP6nf1x1zNUwc6vVafw@mail.gmail.com>
Message-ID: <962ffeb780843553c1f353008459a93f.NginxMailingListEnglish@forum.nginx.org>

> 
> Your feedback is always welcome! Sorry for the
> delay on my side for I
> was enjoying my vocation in the Fuzhou city in the
> southern region of
> China in the last few days :)
> 

Heh, I know what delays are and that was no delay.. Here it is best
weather for bicycling, so i do all the time! ;) One thing on my todo is
to visit china, my cousin lived there past 5-6 years and i've seen many
pictures and videos.. Last month he married his wife there, but i could
not get away :(



> 
> What's your operating system? We used to observe
> similar periodic
> segfaults with exactly the same version of
> ngx_openresty on RedHat
> Enterprise Linux (RHEL) 5.4 in production but now

My good *old* development box is a quad *i686* with deb5 (old)stable gcc
4.3.2-1.1



> all is fine after we
> switched LuaJIT to a more recent git HEAD version
> which includes an
> important work-around for an old libgcc bug
> explained here in the
> lua-l mailing list:
> 
>    
> http://lua-users.org/lists/lua-l/2011-03/msg00576.
> html
> 

Tried both samples and luajit detected it. Anyway thank you for sharing
this, so ill try that next.



> I'll make a new release of ngx_openresty to
> include a new version of
> LuaJIT as well as some other modules' updates like
> ngx_lua's and
> ngx_srcache's. I'll inform you here when I'm done.
> 
> Also, it'll be very appreciated if you can obtain
> a gdb backtrace for
> your segfaults so that we can look at the details
> ;)

Sure, thats no problem. Tomorrow ill attach to it and take a sample for
you. Have also a full dump, as every build has its own /prefix.


> 
> Oh, that's awesome :)
> 

Thanks, but yours is *epic*.



> >
> > Below ill use luasql, instead you can use
> > ngx.location.capture("/some-internal-rds-json")
> and lua-yajl, but this
> > approach ends as soon as you want INSERT
> followed by LAST_INSERT_ID() or
> > other complex queries/transactions (as long as
> libdrizzle has no support
> > for multiple queries, maybe this is not the best
> option to go with, ive
> > tried alternatives and workarounds, but for now
> this is what *I* want).
> >
> 
> The luasql library will certainly block nginx
> worker processes. Please
> dont' use that.
> 

Tbh, youre absolutely right. Realized it before i used this and *I* had
no problems, also do not have much inserts (about 25-40k text each,
stored in a fraction of a second). Tested it also further with luasql
all queries, no problems so far but only with circumventions and all
under pre-calculated conditions with *enough* instances, fast i/o,
caching etc. This is at no will a reliable solution and much likely a
show-stopper! So do *not* try at home eh?!!

In the meantime or maybe forever ill stick to redis only (: Exchanges to
or from intermediates are stressing me, but still better as seeing ruby
eating up resources and doing "nothing" on ngx_passenger/ree.. sinatra,
merb, rails, foo, bla -.-" Redis helped there a lot, but even in my
*experimental* Rack written with D ive done better! ;) But thats true
another story..

As a last resort for mysql i thought of getting an unique id from a
(lua) location which is then used for insertion and afterwards on
success, i.e. for redirects. Or have i missed something, what would you
do? 


> I know libdrizzle is very limited in terms of
> functionality and it's
> *not* maintained very well. We'll start a
> completely new nginx
> upstream modules for mysql named ngx_mysql by
> implementing a
> non-blocking full-fledged highly-optimized mysql
> client from scratch.
> I'd love to see it happen this year ;)
> 
> Another direction we're taking is to introduce the
> "cosocket"
> mechanism into ngx_lua such that one can use a
> *synchronous* Lua
> socket API that is transparently *asynchronous*
> and *non-blocking* on
> the nginx C level. Atop that, we can build
> pure-Lua non-blocking mysql
> client driver and thus putting everything onto the
> Lua land and no
> longer need to mess up the nginx.conf file and
> nginx subrequests. I'd
> also want to see it happen by this Christmas ;)
> 

You mentioned "cosocket" somewhere (github?) already and IMO that would
be really amazing! I'm following "agentzh" and "chaoslawful" on github
and will take a deeper look when i get more time. Will be linking to
your projects when i release something useful (safe) for it.



> 
> The high-level abstraction for ngx_lua is
> excellent! We're taking a
> slightly different approach along the line in our
> web applications:
> 
>    
> http://agentzh.org/misc/nginx/lzsql-manual.html
> 
> We're using a DSL (Domain Specific Language)
> compiler that emits Lua
> code and business logic is mostly implemented in
> this DSL. It's
> targeting our business though, not meant to be
> useful for general web
> application development (yet).
> 

Looks very promising, i seen that page already on my researches and
would love to get my hands on it. I do a lot of research, trying as much
as possible, protocols like XMPP and PSYC and such good things, but
mostly everything apart from PHP, .Net and Java.. :>

> Thanks for sharing!
> 
> -agentzh
> 

Heh. >.< All the thanks belong to you and who else is also behind that
projects! Had so much *fun* with it so far, you wont believe it how it
boosted everything at our groups site, ookeh not *that* much, but enough
to already left ruby (was riding since little rails). And personally it
helped me alot to grok lua, as i needed that for deeper reworks of my
prosody-im fork (another lua project thats awesome, i prefer it over
ejabberd/erl which indeed is both neat feature-wise, but wont let me
change (easily) from mnesia to redis or other nice tricks).


Thanks again!
# HTML

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,179364,195682#msg-195682



From clanherb at gmail.com  Thu May  5 06:14:55 2011
From: clanherb at gmail.com (=?UTF-8?B?5p2o6ZWt?=)
Date: Thu, 5 May 2011 10:14:55 +0800
Subject: question about the context of map directive
In-Reply-To: <20110504203547.GX42265@mdounin.ru>
References: <BANLkTinkf+Dsj3S7bCGd-LwPXnwwne7H6g@mail.gmail.com>
	<20110504203547.GX42265@mdounin.ru>
Message-ID: <BANLkTi=vec8+AS-nWpA44GvYY7CBonh6xA@mail.gmail.com>

Hi Igor and Maxim,

Thanks for the reply!

On Thu, May 5, 2011 at 4:35 AM, Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
>
> On Thu, May 05, 2011 at 02:59:21AM +0800, ?? wrote:
>
> > Hi,
> >
> > The context of map directive is *http* as documented on wiki.nginx.org.
> This
>
> Yes.
>
> > means all the requests will go through the map filter phase.
>
> No.  Map directive creates a variable and only does something when
> (and if) the variable is accessed.  There is no performance
> penalty for processing requests which doesn't use the variable.
>
> > This is not ideal when you have many virtual hosts.
> >
> > For example, we've 50+ virtual hosts. When one of them need map
> variables,
> > the others automatically inherited those settings because there is no way
> to
> > limit it to server context.
> >
> > If this is the situation, can someone know the internals of nginx explain
> > why the map module can't work at server level.
>
> Changing map module to provide data based on per-server configs is
> possible, but obviously will require extra code.
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110505/94632ff8/attachment.html>

From lists at ruby-forum.com  Thu May  5 06:27:50 2011
From: lists at ruby-forum.com (Yanxin Z.)
Date: Thu, 05 May 2011 04:27:50 +0200
Subject: limit connection
In-Reply-To: <3eb40128cef9709ed44491c6d10d7e09@ruby-forum.com>
References: <3eb40128cef9709ed44491c6d10d7e09@ruby-forum.com>
Message-ID: <05c99fd96704fbea856c6e5e906fb1ec@ruby-forum.com>

Thank you for your direction, Chris.

-- 
Posted via http://www.ruby-forum.com/.


From nginx-forum at nginx.us  Thu May  5 06:50:07 2011
From: nginx-forum at nginx.us (stalkercn)
Date: Wed, 04 May 2011 22:50:07 -0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <65b155ed9ec7350b8df7cf617867f95e.NginxMailingListEnglish@forum.nginx.org>

nginx version: nginx/0.8.54
configure arguments: --prefix=/usr/local/etc/nginx --with-cc-opt='-I
/usr/local/include' --with-ld-opt='-L /usr/local/lib'
--conf-path=/usr/local/etc/nginx/nginx.conf
--sbin-path=/usr/local/sbin/nginx --pid-path=/var/run/nginx.pid
--error-log-path=/var/log/nginx-error.log --user=www --group=www
--with-debug
--http-client-body-temp-path=/var/tmp/nginx/client_body_temp
--http-fastcgi-temp-path=/var/tmp/nginx/fastcgi_temp
--http-proxy-temp-path=/var/tmp/nginx/proxy_temp
--http-scgi-temp-path=/var/tmp/nginx/scgi_temp
--http-uwsgi-temp-path=/var/tmp/nginx/uwsgi_temp
--http-log-path=/var/log/nginx-access.log
--add-module=/usr/ports/www/nginx/work/ngx_headers_more_module-0.14
--with-http_flv_module --with-http_gzip_static_module
--with-http_image_filter_module --with-http_stub_status_module
--with-pcre

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,192785,195695#msg-195695



From nginx-forum at nginx.us  Thu May  5 07:00:14 2011
From: nginx-forum at nginx.us (burn)
Date: Wed, 04 May 2011 23:00:14 -0400
Subject: not working without aio
In-Reply-To: <20110504112733.GT42265@mdounin.ru>
References: <20110504112733.GT42265@mdounin.ru>
Message-ID: <3bf60e8194f645c340689fb77744f521.NginxMailingListEnglish@forum.nginx.org>

well, then please consider this a feature request. Having two separate
packages for our server farm is not too convenient.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195522,195696#msg-195696



From agentzh at gmail.com  Thu May  5 07:20:32 2011
From: agentzh at gmail.com (agentzh)
Date: Thu, 5 May 2011 11:20:32 +0800
Subject: Catch ALL requests by LUA-script
In-Reply-To: <962ffeb780843553c1f353008459a93f.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTinB3WNOQ4=pP6nf1x1zNUwc6vVafw@mail.gmail.com>
	<962ffeb780843553c1f353008459a93f.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTi=ENcqEcRiwxZ5gCQDB0o=o-QSmyg@mail.gmail.com>

On Thu, May 5, 2011 at 8:05 AM, HowToMeetLadies <nginx-forum at nginx.us> wrote:
>
> Heh, I know what delays are and that was no delay.. Here it is best
> weather for bicycling, so i do all the time! ;) One thing on my todo is
> to visit china, my cousin lived there past 5-6 years and i've seen many
> pictures and videos.. Last month he married his wife there, but i could
> not get away :(
>

Oh, I'm so happy to hear that ;) Please bring my best wishes to your cousin!

>
> Sure, thats no problem. Tomorrow ill attach to it and take a sample for
> you. Have also a full dump, as every build has its own /prefix.
>

Looking forward to that :)

>
> Tbh, youre absolutely right. Realized it before i used this and *I* had
> no problems, also do not have much inserts (about 25-40k text each,
> stored in a fraction of a second). Tested it also further with luasql
> all queries, no problems so far but only with circumventions and all
> under pre-calculated conditions with *enough* instances, fast i/o,
> caching etc. This is at no will a reliable solution and much likely a
> show-stopper! So do *not* try at home eh?!!
>

Well, it will only cause problems when you have too few nginx workers
and nontrivial number of concurrent requests and your database is
going mad and becoming slow ;) Anyway, it's good for personal use, not
for production apps ;)

> In the meantime or maybe forever ill stick to redis only (: Exchanges to
> or from intermediates are stressing me, but still better as seeing ruby
> eating up resources and doing "nothing" on ngx_passenger/ree.. sinatra,
> merb, rails, foo, bla -.-" Redis helped there a lot, but even in my
> *experimental* Rack written with D ive done better! ;) But thats true
> another story..
>

When you're using ngx_redis2 + lua-redis-parser, please ensure you're
using a tcp connection pool (provided by ngx_http_upstream_keepalive)
and redis pipelining wherever possible. These features will
significantly improve performance. Also, using multiple instance of
redis servers on your multi-core machines also help a lot due to the
sequential processing nature of a single redis server instance.

Also, when you're benchmarking performance, please ensure that your
error log level is high enough to prevent nginx workers spend too much
cycles on flushing the error.log file, which is certainly very
expensive :)

> As a last resort for mysql i thought of getting an unique id from a
> (lua) location which is then used for insertion and afterwards on
> success, i.e. for redirects. Or have i missed something, what would you
> do?
>

You can certainly generate a globally unique id by means of pure Lua
or by accessing redis, though that's a bit silly ;)

If all you want is to get LAST_INSERT_ID, then ngx_drizzle already
returns that automatically for you when you're doing a SQL insert
query. Consider the following sample nginx.conf snippet:

    location /test {
        echo_location /mysql "drop table if exists foo";
        echo;
        echo_location /mysql "create table foo (id serial not null,
primary key (id), val real);";
        echo;
        echo_location /mysql "insert into foo (val) values (3.1415926);";
        echo;
        echo_location /mysql "select * from foo;";
        echo;
    }
    location /mysql {
        drizzle_pass backend;
        drizzle_module_header off;
        drizzle_query $query_string;
        rds_json on;
    }

Then GET /test gives the following outputs:

    {"errcode":0}
    {"errcode":0}
    {"errcode":0,"insert_id":1,"affected_rows":1}
    [{"id":1,"val":3.1415926}]

Have you noticed the "insert_id" field in the 3rd JSON response? ;)

Full transaction support will land into ngx_drizzle (maybe
ngx_postgres as well). It will be implemented by locking and tagging
database connections in the pool. But there's no ETA yet for that.

>
> You mentioned "cosocket" somewhere (github?) already and IMO that would
> be really amazing! I'm following "agentzh" and "chaoslawful" on github
> and will take a deeper look when i get more time. Will be linking to
> your projects when i release something useful (safe) for it.
>

Oh, I'm feeling honored :)

>
> Looks very promising, i seen that page already on my researches and
> would love to get my hands on it. I do a lot of research, trying as much
> as possible, protocols like XMPP and PSYC and such good things, but
> mostly everything apart from PHP, .Net and Java.. :>
>

LOL

>
> Heh. >.< All the thanks belong to you and who else is also behind that
> projects! Had so much *fun* with it so far, you wont believe it how it
> boosted everything at our groups site, ookeh not *that* much, but enough
> to already left ruby (was riding since little rails). And personally it
> helped me alot to grok lua, as i needed that for deeper reworks of my
> prosody-im fork (another lua project thats awesome, i prefer it over
> ejabberd/erl which indeed is both neat feature-wise, but wont let me
> change (easily) from mnesia to redis or other nice tricks).
>

ngx_openresty still has many rough corners and has a long way to go.
We'll surely try to make it better and better unfailingly :)

Happy hacking!

-agentzh


From chaoslawful at gmail.com  Thu May  5 07:36:57 2011
From: chaoslawful at gmail.com (Chaos Wang)
Date: Thu, 05 May 2011 11:36:57 +0800
Subject: Lua pcre inside nginx
In-Reply-To: <C9E71B45.DD104%Brian.Akins@turner.com>
References: <C9E71B45.DD104%Brian.Akins@turner.com>
Message-ID: <4DC21B59.4010309@gmail.com>

NginX hooked pcre memory alloc/free functions pcre_malloc/free to use 
its memory pool, in order to work around its memory leaking problem. As 
a side effect, it reset this pool to NULL after every usage in core 
without considering possible external references. If you don't set pcre 
memory hooks to your own implementation, you will surely get the error.

ngx_lua has fixed this problem, and can use pcre related extensions 
normally after v0.1.6rc5.

Ref:
* src/core/ngx_regex.c in NginX sources
* ngx_http_lua_patch.c in ngx_lua sources

On 2011?05?05? 03:03, Akins, Brian wrote:
> I wanted to use pcre inside a lua script using rex_pcre.  However, I always
> get the error:
>
> failed to get memory (pattern offset: 15)
>
> The snippet of code always works in a standalone script.
>
> Ideas?
>



From nginx-forum at nginx.us  Thu May  5 08:11:30 2011
From: nginx-forum at nginx.us (AndrewH)
Date: Thu, 05 May 2011 00:11:30 -0400
Subject: Pre-process request via upstream then forward to proxy.
Message-ID: <d5e929b747592747889b10590c820932.NginxMailingListEnglish@forum.nginx.org>

As part of processing a request, I need to first send and receive a
response to/from an upstream to authenticate some credentials in the
request.  Once this has been done, the original request needs to be
proxied to another upstream.

According to Evan Miller's tutorial, the only valid location for
spawning sub-requests is in an output filter ie. to post-process a
response.  So it appears that this approach is useful for massaging a
reponse but not valid for actually performing pre-processing on a
request.

Does anyone know what the sanctioned way of going about this in NGINX
is?  Or which existing modules/examples implement similar
functionality?

- Andrew

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195702,195702#msg-195702



From Richard.Kearsley at m247.com  Thu May  5 09:05:37 2011
From: Richard.Kearsley at m247.com (Richard Kearsley)
Date: Thu, 5 May 2011 05:05:37 +0000
Subject: aio info
Message-ID: <EFC5A19DBF3A2E489904733B467D990B0EC5BE@office-server2.m247.local>

Hi
A lot of talk about AIO recently - I actually didn't know it existed before, so I am trying it on freebsd8

# sysctl -a | grep aio
kern.features.aio: 1
vfs.aio.max_buf_aio: 16
vfs.aio.max_aio_queue_per_proc: 256
vfs.aio.max_aio_per_proc: 32
vfs.aio.unloadable: 0
vfs.aio.aiod_lifetime: 30000
vfs.aio.aiod_timeout: 10000
vfs.aio.num_buf_aio: 0
vfs.aio.num_queue_count: 0
vfs.aio.max_aio_queue: 1024
vfs.aio.target_aio_procs: 4
vfs.aio.num_aio_procs: 4
vfs.aio.max_aio_procs: 32

Could anyone comment what these mean and how to go about tweaking for best performance (should any be increased?)
Particularly.. vfs.aio.num_queue_count - does this monitor the amount of io waiting?

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110505/63fc27da/attachment.html>

From agentzh at gmail.com  Thu May  5 10:11:32 2011
From: agentzh at gmail.com (agentzh)
Date: Thu, 5 May 2011 14:11:32 +0800
Subject: Pre-process request via upstream then forward to proxy.
In-Reply-To: <d5e929b747592747889b10590c820932.NginxMailingListEnglish@forum.nginx.org>
References: <d5e929b747592747889b10590c820932.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTikz=4J+MuG5UYrRRo+Vh_rnUSdwoQ@mail.gmail.com>

On Thu, May 5, 2011 at 12:11 PM, AndrewH <nginx-forum at nginx.us> wrote:
> As part of processing a request, I need to first send and receive a
> response to/from an upstream to authenticate some credentials in the
> request. ?Once this has been done, the original request needs to be
> proxied to another upstream.
>

Your task can be trivially done in pure Lua by using the ngx_lua
module. Specifically, by using the ngx.location.capture() interface
for Lua to issue your subrequests and wait for the responses.

See http://github.com/chaoslawful/lua-nginx-module for the full documentation.

You can surely code it up in pure C by looking at how ngx_lua does on
the C level. But it's not recommended because it's error prone and the
performance gain is small.

> According to Evan Miller's tutorial, the only valid location for
> spawning sub-requests is in an output filter ie. to post-process a
> response.

Nope, you can surely create subrequests in any of your rewrite,
access, and content phase handlers. That part of Evan Miller's
tutorial is somehow misleading, and keep in mind that Evan stated like
this: "as far as I know..." So he is not to blame ;)

> Does anyone know what the sanctioned way of going about this in NGINX
> is? ?Or which existing modules/examples implement similar
> functionality?
>

See the echo_subrequest and echo_location directives provided by ngx_echo:

    http://github.com/agentzh/echo-nginx-module

But their (synchronous but non-blocking) subrequest implementation is
not as correct as ngx_lua's.

The ngx_auth_request module by Maxim Dounin is an example of issuing
subrequests in an access phase handler and my ngx_srcache module is
another example for this:

    http://github.com/agentzh/srcache-nginx-module

The ngx_eval module is an example of issuing subrequests in a rewrite
phase handler:

    http://www.grid.net.ru/nginx/eval.en.html

Happy subrequesting!

-agentzh


From mdounin at mdounin.ru  Thu May  5 10:22:06 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 5 May 2011 10:22:06 +0400
Subject: no server response in https
In-Reply-To: <f2292582eeb12229fb3a95a01b37b61e.NginxMailingListEnglish@forum.nginx.org>
References: <f2292582eeb12229fb3a95a01b37b61e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110505062206.GY42265@mdounin.ru>

Hello!

On Tue, May 03, 2011 at 01:42:42PM -0400, art.wu wrote:

> I'm running 0.8.53 on QNX Neutrino 6.5.0 with SSL on and the server
> doesn't respond to the client. Instead it times out, as below. It does
> work properly with SSL off. Can you provide any suggestions? Thanks.

Yep, this is general bug affecting systems with 32-bit off_t.  
Attached patch fixes it.

Please note that using 64-bit off_t may be a good thing anyway (2G 
limit is two low today even on embedded platforms...), and it 
looks like QNX (at least 6.5.0) supports -D_FILE_OFFSET_BITS=64 
(though nginx doesn't know and doesn't try it on QNX).  Try using 

./configure --with-cc-opt="-D_FILE_OFFSET_BITS=64" ...

it should resolve the issue even without the patch.

Please also note that there are other problems with QNX 6 due to 
time_t type being unsigned there.  GCC will refuse to build nginx 
with -Werror (as used by nginx) - and this is the right thing.  
I'm looking into this issue, but it will require a bit more work.

Maxim Dounin
-------------- next part --------------
# HG changeset patch
# User Maxim Dounin <mdounin at mdounin.ru>
# Date 1304575111 -14400
# Node ID 0446d18df946f8cfae3f623d3335722d7c4489a6
# Parent  9b9e61f1b5004e0c5a2d56372e682c2a1e053e34
Bugfix: https wasn't working on systems with 32-bit off_t.

Due to off_t being signed NGX_MAX_UINT32_VALUE will overflow off_t while
calculating limit in ngx_ssl_send_chain() on systems with 32-bit time_t,
and as a result ngx_ssl_send_chain() won't be able to send anything.

Introduce NGX_MAX_INT32_VALUE and use it there instead.

diff --git a/src/core/ngx_config.h b/src/core/ngx_config.h
--- a/src/core/ngx_config.h
+++ b/src/core/ngx_config.h
@@ -127,5 +127,6 @@ typedef intptr_t        ngx_flag_t;
 #define NGX_MAX_UINT32_VALUE  (uint32_t) 0xffffffff
 #endif
 
+#define NGX_MAX_INT32_VALUE   (int32_t) 2147483647
 
 #endif /* _NGX_CONFIG_H_INCLUDED_ */
diff --git a/src/event/ngx_event_openssl.c b/src/event/ngx_event_openssl.c
--- a/src/event/ngx_event_openssl.c
+++ b/src/event/ngx_event_openssl.c
@@ -957,10 +957,10 @@ ngx_ssl_send_chain(ngx_connection_t *c, 
     }
 
 
-    /* the maximum limit size is the maximum uint32_t value - the page size */
-
-    if (limit == 0 || limit > (off_t) (NGX_MAX_UINT32_VALUE - ngx_pagesize)) {
-        limit = NGX_MAX_UINT32_VALUE - ngx_pagesize;
+    /* the maximum limit size is the maximum int32_t value - the page size */
+
+    if (limit == 0 || limit > (off_t) (NGX_MAX_INT32_VALUE - ngx_pagesize)) {
+        limit = NGX_MAX_INT32_VALUE - ngx_pagesize;
     }
 
     buf = c->ssl->buf;

From mdounin at mdounin.ru  Thu May  5 10:31:10 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 5 May 2011 10:31:10 +0400
Subject: Pre-process request via upstream then forward to proxy.
In-Reply-To: <d5e929b747592747889b10590c820932.NginxMailingListEnglish@forum.nginx.org>
References: <d5e929b747592747889b10590c820932.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110505063110.GZ42265@mdounin.ru>

Hello!

On Thu, May 05, 2011 at 12:11:30AM -0400, AndrewH wrote:

> As part of processing a request, I need to first send and receive a
> response to/from an upstream to authenticate some credentials in the
> request.  Once this has been done, the original request needs to be
> proxied to another upstream.
> 
> According to Evan Miller's tutorial, the only valid location for
> spawning sub-requests is in an output filter ie. to post-process a
> response.  So it appears that this approach is useful for massaging a
> reponse but not valid for actually performing pre-processing on a
> request.

Evan Miller's claim was correct before 0.7.25.  In recent versions 
you may issue subrequests almost everywhere.  There are some 
caveats though - it's up to you to take appropriate measures to 
ensure subrequest response won't appear on client's connection.

> Does anyone know what the sanctioned way of going about this in NGINX
> is?  Or which existing modules/examples implement similar
> functionality?

Take a look at:

http://grid.net.ru/nginx/eval.en.html
http://mdounin.ru/hg/ngx_http_auth_request_module/

Maxim Dounin


From igor at sysoev.ru  Thu May  5 11:12:12 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Thu, 5 May 2011 11:12:12 +0400
Subject: aio info
In-Reply-To: <EFC5A19DBF3A2E489904733B467D990B0EC5BE@office-server2.m247.local>
References: <EFC5A19DBF3A2E489904733B467D990B0EC5BE@office-server2.m247.local>
Message-ID: <20110505071212.GA36559@sysoev.ru>

On Thu, May 05, 2011 at 05:05:37AM +0000, Richard Kearsley wrote:
> Hi
> A lot of talk about AIO recently - I actually didn't know it existed before, so I am trying it on freebsd8
> 
> # sysctl -a | grep aio
> kern.features.aio: 1
> vfs.aio.max_buf_aio: 16
> vfs.aio.max_aio_queue_per_proc: 256
> vfs.aio.max_aio_per_proc: 32
> vfs.aio.unloadable: 0
> vfs.aio.aiod_lifetime: 30000
> vfs.aio.aiod_timeout: 10000
> vfs.aio.num_buf_aio: 0
> vfs.aio.num_queue_count: 0
> vfs.aio.max_aio_queue: 1024
> vfs.aio.target_aio_procs: 4
> vfs.aio.num_aio_procs: 4
> vfs.aio.max_aio_procs: 32
> 
> Could anyone comment what these mean and how to go about tweaking for best performance (should any be increased?)

$ sysctl -d vfs.aio
vfs.aio: Async IO management
vfs.aio.max_buf_aio: Maximum buf aio requests per process (stored in the process)
vfs.aio.max_aio_queue_per_proc: Maximum queued aio requests per process (stored in the process)
vfs.aio.max_aio_per_proc: Maximum active aio requests per process (stored in the process)
vfs.aio.unloadable: Allow unload of aio (not recommended)
vfs.aio.aiod_lifetime: Maximum lifetime for idle aiod
vfs.aio.aiod_timeout: Timeout value for synchronous aio operations
vfs.aio.num_buf_aio: Number of aio requests presently handled by the buf subsystem
vfs.aio.num_queue_count: Number of queued aio requests
vfs.aio.max_aio_queue: Maximum number of aio requests to queue, globally
vfs.aio.target_aio_procs: Preferred number of ready kernel threads for async IO
vfs.aio.num_aio_procs: Number of presently active kernel threads for async IO
vfs.aio.max_aio_procs: Maximum number of kernel threads to use for handling async IO 

There are two types AIO operations inside FreeBSD kernel - raw file
AIO (buf aio) and usual file AIO that uses in-kernel threads/processes.
nginx uses only usual file AIO. Relevant sysctls are:

vfs.aio.max_aio_queue: 1024
vfs.aio.max_aio_queue_per_proc: 256
vfs.aio.max_aio_per_proc: 32
vfs.aio.num_queue_count: 0

vfs.aio.max_aio_procs: 32
vfs.aio.target_aio_procs: 4
vfs.aio.num_aio_procs: 4

vfs.aio.aiod_lifetime: 30000
vfs.aio.aiod_timeout: 10000

You may want to increase

vfs.aio.max_aio_procs
vfs.aio.max_aio_queue
vfs.aio.max_aio_queue_per_proc
vfs.aio.max_aio_per_proc

> Particularly.. vfs.aio.num_queue_count - does this monitor the amount of io waiting?

Yes.


-- 
Igor Sysoev


From igor at sysoev.ru  Thu May  5 11:34:45 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Thu, 5 May 2011 11:34:45 +0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <65b155ed9ec7350b8df7cf617867f95e.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
	<65b155ed9ec7350b8df7cf617867f95e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110505073445.GC36559@sysoev.ru>

On Wed, May 04, 2011 at 10:50:07PM -0400, stalkercn wrote:
> nginx version: nginx/0.8.54
> configure arguments: --prefix=/usr/local/etc/nginx --with-cc-opt='-I
> /usr/local/include' --with-ld-opt='-L /usr/local/lib'
> --conf-path=/usr/local/etc/nginx/nginx.conf
> --sbin-path=/usr/local/sbin/nginx --pid-path=/var/run/nginx.pid
> --error-log-path=/var/log/nginx-error.log --user=www --group=www
> --with-debug
> --http-client-body-temp-path=/var/tmp/nginx/client_body_temp
> --http-fastcgi-temp-path=/var/tmp/nginx/fastcgi_temp
> --http-proxy-temp-path=/var/tmp/nginx/proxy_temp
> --http-scgi-temp-path=/var/tmp/nginx/scgi_temp
> --http-uwsgi-temp-path=/var/tmp/nginx/uwsgi_temp
> --http-log-path=/var/log/nginx-access.log
> --add-module=/usr/ports/www/nginx/work/ngx_headers_more_module-0.14
> --with-http_flv_module --with-http_gzip_static_module
> --with-http_image_filter_module --with-http_stub_status_module
> --with-pcre

There was a bug when nginx was built --without-http_cache, but this is
not your case. Could you try nginx-1.0.0 without third-party module
ngx_headers_more_module ?


-- 
Igor Sysoev


From CBoyce at mersofmich.com  Thu May  5 03:00:11 2011
From: CBoyce at mersofmich.com (Collin Boyce)
Date: Wed, 4 May 2011 23:00:11 +0000
Subject: Exchange 2010 rpc ??
Message-ID: <2B45E42FC139C849A48677B3750F61810228FA@EMAIL-01.mersdomain1>

We are not able to get nginx to reverse procy RPC to exchange 2010.

We get repeating error messages in the log:
x.x.x.x - - [04/May/2011:08:40:43 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:40:43 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:41:04 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:41:04 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:41:38 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:41:38 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:41:59 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:41:59 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:42:38 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:42:38 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:42:59 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:42:59 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:43:38 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:43:38 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:44:00 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:44:00 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:44:38 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:44:38 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:44:59 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:44:59 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:49:39 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:49:39 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:50:00 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:50:00 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:54:39 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:54:39 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:55:00 -0400] "-" 400 173 "-" "-"
x.x.x.x - - [04/May/2011:08:55:00 -0400] "-" 400 173 "-" "-"


The testexchangeconnectivity tool bombs with the following error:



Config lines:

        location /rpc  { proxy_pass https://mail.local/rpc; }






Checking the IIS configuration for client certificate authentication.



The test passed with some warnings encountered. Please expand the additional details.



[Description: https://www.testexchangeconnectivity.com/Images/Minus.gif]

Additional Details



Client certificate authentication couldn't be determined because an unexpected failure occurred. WinHttpSendRequest failed with error 12152.



[Description: https://www.testexchangeconnectivity.com/Images/Error.png]

Testing HTTP Authentication Methods for URL https://mail.local/rpc/rpcproxy.dll.



The HTTP authentication test failed.



[Description: https://www.testexchangeconnectivity.com/Images/Minus.gif]

Additional Details



Exception details:
Message: The server committed a protocol violation. Section=ResponseStatusLine
Type: System.Net.WebException
Stack trace:
at System.Net.HttpWebRequest.GetResponse()
at Microsoft.Exchange.Tools.ExRca.Tests.HttpAuthMethodsTest.GetSupportedHttpAuthMethods()
at Microsoft.Exchange.Tools.ExRca.Tests.HttpAuthMethodsTest.PerformTestReally()







--

CONFIDENTIALITY NOTICE:

This email and any attachments are for the sole use of the intended recipient(s) and may contain information that is confidential and protected from disclosure under the law, including attorney-client communications. Any unauthorized review, use, disclosure, or distribution is prohibited.  If you are not the intended recipient, please contact the sender by reply email and delete/destroy all copies of the original message and attachments.

DISCLAIMER:

This email may contain a summary description of the Municipal Employees? Retirement System of Michigan benefits, costs, rates, valuations, or other calculations, policies or procedures. MERS has made every effort to ensure, but does not guarantee that the information provided is accurate and up to date. Where this email conflicts with the relevant MERS Plan Document, the MERS Plan Document controls.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110504/bba1e6e5/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.gif
Type: image/gif
Size: 853 bytes
Desc: image001.gif
URL: <http://nginx.org/pipermail/nginx/attachments/20110504/bba1e6e5/attachment.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image002.png
Type: image/png
Size: 911 bytes
Desc: image002.png
URL: <http://nginx.org/pipermail/nginx/attachments/20110504/bba1e6e5/attachment.png>

From pasik at iki.fi  Thu May  5 17:23:14 2011
From: pasik at iki.fi (Pasi =?iso-8859-1?Q?K=E4rkk=E4inen?=)
Date: Thu, 5 May 2011 16:23:14 +0300
Subject: nginx proxying do not locally save/buffer uploaded files?
In-Reply-To: <3dfd6d89a78e3ee124ed58524312dada.NginxMailingListEnglish@forum.nginx.org>
References: <20110504214005.GM32595@reaktio.net>
	<3dfd6d89a78e3ee124ed58524312dada.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110505132314.GP32595@reaktio.net>

On Wed, May 04, 2011 at 05:56:43PM -0400, jjjx128 wrote:
> Pasi K?rkk?inen Wrote:
> -------------------------------------------------------
> > in tcp mode you can't filter the requests etc.. I
> > need to do that aswell.
> 
> There are quite a few load balancers and proxies out there. I'm sure you
> can find something useful.
> 

I've checked and/or used most of them :)
 
> > Maybe I need to take a look at the code and try to
> > figure out
> > if direct upload streaming could be implemented in
> > nginx..
> 
> It is possible. So, good luck I guess.
> 

Thanks. Are you familiar with nginx code? 
If yes, feel free to give advises what to avoid,
and what to look for.

(I haven't checked nginx code earlier, but I know C)

-- Pasi



From nginx-forum at nginx.us  Thu May  5 19:25:19 2011
From: nginx-forum at nginx.us (art.wu)
Date: Thu, 05 May 2011 11:25:19 -0400
Subject: no server response in https
In-Reply-To: <f2292582eeb12229fb3a95a01b37b61e.NginxMailingListEnglish@forum.nginx.org>
References: <f2292582eeb12229fb3a95a01b37b61e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <abee39713bb07afdb53b317067e49eb5.NginxMailingListEnglish@forum.nginx.org>

The configure option solved the problem. Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195329,195791#msg-195791



From lists at ruby-forum.com  Thu May  5 19:52:43 2011
From: lists at ruby-forum.com (Yanxin Z.)
Date: Thu, 05 May 2011 17:52:43 +0200
Subject: memcache issue
Message-ID: <eb9a7f3590d86beb5fedd38d4ceb695d@ruby-forum.com>

Hello,
I am using PHP to connect to memcache.

I connect memcache service in the PHP code as follows:
 $memcache = new Memcache;
 $memcache->connect('127.0.0.1', 11211); //connect to memcached server


Then when I connect to NGX server, I got 500 or 502 error.

However, if I run the memcache using php test.php directly, I can
connect to memcache server successfully.

I am using php-fpm.

Does anyone encounter the similar issue before?

Thanks,
Yanxin

-- 
Posted via http://www.ruby-forum.com/.


From igor at sysoev.ru  Thu May  5 20:01:29 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Thu, 5 May 2011 20:01:29 +0400
Subject: memcache issue
In-Reply-To: <eb9a7f3590d86beb5fedd38d4ceb695d@ruby-forum.com>
References: <eb9a7f3590d86beb5fedd38d4ceb695d@ruby-forum.com>
Message-ID: <66EB54CA-3655-4BF3-BA6B-5FAF2910D8A1@sysoev.ru>

What is in error_log ?

--
Igor Sysoev

05.05.2011, ? 19:52, "Yanxin Z." <lists at ruby-forum.com> ???????(?):

> Hello,
> I am using PHP to connect to memcache.
> 
> I connect memcache service in the PHP code as follows:
> $memcache = new Memcache;
> $memcache->connect('127.0.0.1', 11211); //connect to memcached server
> 
> 
> Then when I connect to NGX server, I got 500 or 502 error.
> 
> However, if I run the memcache using php test.php directly, I can
> connect to memcache server successfully.
> 
> I am using php-fpm.
> 
> Does anyone encounter the similar issue before?
> 
> Thanks,
> Yanxin
> 
> -- 
> Posted via http://www.ruby-forum.com/.
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx


From lists at ruby-forum.com  Thu May  5 20:10:27 2011
From: lists at ruby-forum.com (Yanxin Z.)
Date: Thu, 05 May 2011 18:10:27 +0200
Subject: memcache issue
In-Reply-To: <eb9a7f3590d86beb5fedd38d4ceb695d@ruby-forum.com>
References: <eb9a7f3590d86beb5fedd38d4ceb695d@ruby-forum.com>
Message-ID: <2f77714110eed19c0f3cc49c50d4cc3b@ruby-forum.com>

Thank you for your help, Igor.

The log is


1815426 2011/05/05 09:05:09 [notice] 13424#0: *15 
"/api/1.0/(.*)/upload_internal" matches "/api/1.0/ 
web/upload_internal", client: 10.1.4.250, server: localhost, request: 
"POST /api/1.0/web/upload_internal HTTP/1.1", host: "10.1.4.243"
1815427 2011/05/05 09:05:09 [notice] 13424#0: *15 rewritten data: 
"/api/1.0/web/upload.php", args: "",         client: 10.1.4.250, server: 
localhost, request: "POST /api/1.0/web/upload_internal HTTP/1.1", 
host: "10.1.4.243"
1815428 2011/05/05 09:05:09 [error] 13424#0: *15 recv() failed (104: 
Connection reset by peer) while reading response header from upstream, 
client: 10.1.4.250, server: localhost, request: "POST / 
api/1.0/web/upload_internal HTTP/1.1", upstream: 
"fastcgi://127.0.0.1:9000", host: "10.1.4.243"
1815429 2011/05/05 09:05:09 [info] 13424#0: *15 client 10.1.4.250 closed 
keepalive connection

-- 
Posted via http://www.ruby-forum.com/.


From lists at ruby-forum.com  Thu May  5 20:28:02 2011
From: lists at ruby-forum.com (Yanxin Z.)
Date: Thu, 05 May 2011 18:28:02 +0200
Subject: memcache issue
In-Reply-To: <eb9a7f3590d86beb5fedd38d4ceb695d@ruby-forum.com>
References: <eb9a7f3590d86beb5fedd38d4ceb695d@ruby-forum.com>
Message-ID: <e870194b26a435f568f5895f079033b8@ruby-forum.com>

Hello Igor,
Thank you for your direction.

I searched for  "nginx 104:  Connection reset by peer"

In the forum
http://forum.slicehost.com/comments.php?DiscussionID=433

Someone suggest to restart fastcgi.

I did restart php-fpm then, memcached is working now.

Hopefully my post is helpful to other members.
Yanxin

-- 
Posted via http://www.ruby-forum.com/.


From nginx-forum at nginx.us  Thu May  5 20:29:00 2011
From: nginx-forum at nginx.us (HowToMeetLadies)
Date: Thu, 05 May 2011 12:29:00 -0400
Subject: Catch ALL requests by LUA-script
In-Reply-To: <962ffeb780843553c1f353008459a93f.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTinB3WNOQ4=pP6nf1x1zNUwc6vVafw@mail.gmail.com>
	<962ffeb780843553c1f353008459a93f.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <43a20ba24767aaf66b89c49525531ecd.NginxMailingListEnglish@forum.nginx.org>

> Oh, I'm so happy to hear that ;) Please bring my
> best wishes to your cousin!
> 
> >
> > Sure, thats no problem. Tomorrow ill attach to
> it and take a sample for
> > you. Have also a full dump, as every build has
> its own /prefix.
> >
> 
> Looking forward to that :)

Will do both as soon i'll arrive at home later this day!


> 
> When you're using ngx_redis2 + lua-redis-parser,
> please ensure you're
> using a tcp connection pool (provided by
> ngx_http_upstream_keepalive)
> and redis pipelining wherever possible. These
> features will
> significantly improve performance. Also, using
> multiple instance of
> redis servers on your multi-core machines also
> help a lot due to the
> sequential processing nature of a single redis
> server instance.
> 
> Also, when you're benchmarking performance, please
> ensure that your
> error log level is high enough to prevent nginx
> workers spend too much
> cycles on flushing the error.log file, which is
> certainly very
> expensive :)
> 

Thank you for this, my conf applied already to this, but maybe you
should tell about that in your docs too. (:



> > As a last resort for mysql i thought of getting
> an unique id from a
> > (lua) location which is then used for insertion
> and afterwards on
> > success, i.e. for redirects. Or have i missed
> something, what would you
> > do?
> >
> 
> You can certainly generate a globally unique id by
> means of pure Lua
> or by accessing redis, though that's a bit silly
> ;)
> 
> If all you want is to get LAST_INSERT_ID, then
> ngx_drizzle already
> returns that automatically for you when you're
> doing a SQL insert
> query. Consider the following sample nginx.conf
> snippet:
> 
>     location /test {
>         echo_location /mysql "drop table if exists
> foo";
>         echo;
>         echo_location /mysql "create table foo (id
> serial not null,
> primary key (id), val real);";
>         echo;
>         echo_location /mysql "insert into foo
> (val) values (3.1415926);";
>         echo;
>         echo_location /mysql "select * from foo;";
>         echo;
>     }
>     location /mysql {
>         drizzle_pass backend;
>         drizzle_module_header off;
>         drizzle_query $query_string;
>         rds_json on;
>     }
> 
> Then GET /test gives the following outputs:
> 
>     {"errcode":0}
>     {"errcode":0}
>     {"errcode":0,"insert_id":1,"affected_rows":1}
>     [{"id":1,"val":3.1415926}]
> 
> Have you noticed the "insert_id" field in the 3rd
> JSON response? ;)
> 

Unbelievable! :D Ive seen "insert id" header in ngx_drizzle, but hadnt
looked further at inserts with rds_json (damn stupid me, i should write
an article about that).

> 
> LOL
> 
I have a php-fpm install for exactly 3 apps which are our
webmail/cms/board, because php-users are not aware of their risks ive
disabled it apart from that. NCARS rulez, but im replacing it already ->
xmpp (psyc) + web frontend.

I used techniques like MEF and such in .Net, but imho its desktop-only..
One can do amazing things with it, also without using client-side
runtime, but if you're not bound to that platform, why would you do?

Its good to have varnish, ha and/or simple nginx in front of such
servers/frameworks, but now nginx+lua is enough, its so tiny and fast..

> 
> Happy hacking!
> 

Greetz (;
# HTML

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,179364,195814#msg-195814



From nginx-forum at nginx.us  Thu May  5 22:59:24 2011
From: nginx-forum at nginx.us (Jim Ohlstein)
Date: Thu, 05 May 2011 14:59:24 -0400
Subject: Nginx Forum now accessible via IPv6
Message-ID: <18c2af52be92de23fc8c029b1e344080.NginxMailingListEnglish@forum.nginx.org>

Generally I prefer to keep forum business on the forum but given the
current state of IPv4 addresses and the upcoming World IPv6 Day on June
8, 2011 - see http://isoc.org/wp/worldipv6day/ - I am pleased to
announce that the forum now has "native" IPv6 support.

Most people who have an IPv4/IPv6 dual stack either natively from their
ISP or by using a tunnel, and is using a modern browser, will likely be
connecting via IPv6.

Please let me know if there any connectivity issues, either via reply or
directly at jim [at] ohlste [dot] in.

Many Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195868,195868#msg-195868



From nginx-forum at nginx.us  Fri May  6 06:45:21 2011
From: nginx-forum at nginx.us (stalkercn)
Date: Thu, 05 May 2011 22:45:21 -0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <80af9b27411dbe556335e96bd8912214.NginxMailingListEnglish@forum.nginx.org>

nginx: nginx version: nginx/1.0.1
nginx: built by gcc 4.2.1 20070719  [FreeBSD]
nginx: configure arguments: --prefix=/usr/local/etc/nginx
--with-cc-opt='-I /usr/local/include' --with-ld-opt='-L /usr/local/lib'
--conf-path=/usr/local/etc/nginx/nginx.conf
--sbin-path=/usr/local/sbin/nginx --pid-path=/var/run/nginx.pid
--error-log-path=/var/log/nginx.log --user=www --group=www --with-debug
--http-client-body-temp-path=/var/tmp/nginx/client_body_temp
--http-fastcgi-temp-path=/var/tmp/nginx/fastcgi_temp
--http-proxy-temp-path=/var/tmp/nginx/proxy_temp
--http-scgi-temp-path=/var/tmp/nginx/scgi_temp
--http-uwsgi-temp-path=/var/tmp/nginx/uwsgi_temp
--http-log-path=/var/log/nginx-access.log --with-http_gzip_static_module
--with-http_stub_status_module --with-pcre

I decrease the fastcgi buffer size and increase the amount of buffers:
            fastcgi_buffer_size          64k;
            fastcgi_buffers           64 64k;

but a 'too big header' error still apears:

 [debug] upstream split a header line in FastCGI records
 [error]  upstream sent too big header while reading response header
from upstream

btw,php is working with Zend-Optimizer,whether this condition can make
the headers bigger than its real size?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,192785,195960#msg-195960



From agentzh at gmail.com  Fri May  6 08:26:41 2011
From: agentzh at gmail.com (agentzh)
Date: Fri, 6 May 2011 12:26:41 +0800
Subject: Catch ALL requests by LUA-script
In-Reply-To: <43a20ba24767aaf66b89c49525531ecd.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTinB3WNOQ4=pP6nf1x1zNUwc6vVafw@mail.gmail.com>
	<962ffeb780843553c1f353008459a93f.NginxMailingListEnglish@forum.nginx.org>
	<43a20ba24767aaf66b89c49525531ecd.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTinoYw0=ZZHxbGAfpbeyZqdz57gFUg@mail.gmail.com>

On Fri, May 6, 2011 at 12:29 AM, HowToMeetLadies <nginx-forum at nginx.us> wrote:
>>
>> Also, when you're benchmarking performance, please
>> ensure that your
>> error log level is high enough to prevent nginx
>> workers spend too much
>> cycles on flushing the error.log file, which is
>> certainly very
>> expensive :)
>>
>
> Thank you for this, my conf applied already to this, but maybe you
> should tell about that in your docs too. (:
>

Just added to ngx_redis2's README ;)

>> Then GET /test gives the following outputs:
>>
>> ? ? {"errcode":0}
>> ? ? {"errcode":0}
>> ? ? {"errcode":0,"insert_id":1,"affected_rows":1}
>> ? ? [{"id":1,"val":3.1415926}]
>>
>> Have you noticed the "insert_id" field in the 3rd
>> JSON response? ;)
>>
>
> Unbelievable! :D Ive seen "insert id" header in ngx_drizzle, but hadnt
> looked further at inserts with rds_json (damn stupid me, i should write
> an article about that).

Just added this to ngx_drizzle's README too :) My bad, it's indeed not
that obvious for new users ;)

Thanks!
-agentzh


From mcabral84 at bol.com.br  Fri May  6 05:33:02 2011
From: mcabral84 at bol.com.br (Marcelo Cabral)
Date: Thu, 5 May 2011 22:33:02 -0300
Subject: FastCGI server status
Message-ID: <BANLkTi=aJZWZ4FfTVe8kuszn3yni0ot75g@mail.gmail.com>

Hello,

I was wondering if I could monitor FastCGI responses (timers, success,
failures, round robin stats, etc...) with nginx?
Any hints on how I could get this info?

Thanks
-Daniel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110505/8da1037d/attachment.html>

From igor at sysoev.ru  Fri May  6 08:42:32 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 08:42:32 +0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <80af9b27411dbe556335e96bd8912214.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
	<80af9b27411dbe556335e96bd8912214.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110506044232.GE85186@sysoev.ru>

On Thu, May 05, 2011 at 10:45:21PM -0400, stalkercn wrote:
> nginx: nginx version: nginx/1.0.1
> nginx: built by gcc 4.2.1 20070719  [FreeBSD]
> nginx: configure arguments: --prefix=/usr/local/etc/nginx
> --with-cc-opt='-I /usr/local/include' --with-ld-opt='-L /usr/local/lib'
> --conf-path=/usr/local/etc/nginx/nginx.conf
> --sbin-path=/usr/local/sbin/nginx --pid-path=/var/run/nginx.pid
> --error-log-path=/var/log/nginx.log --user=www --group=www --with-debug
> --http-client-body-temp-path=/var/tmp/nginx/client_body_temp
> --http-fastcgi-temp-path=/var/tmp/nginx/fastcgi_temp
> --http-proxy-temp-path=/var/tmp/nginx/proxy_temp
> --http-scgi-temp-path=/var/tmp/nginx/scgi_temp
> --http-uwsgi-temp-path=/var/tmp/nginx/uwsgi_temp
> --http-log-path=/var/log/nginx-access.log --with-http_gzip_static_module
> --with-http_stub_status_module --with-pcre
> 
> I decrease the fastcgi buffer size and increase the amount of buffers:
>             fastcgi_buffer_size          64k;
>             fastcgi_buffers           64 64k;
> 
> but a 'too big header' error still apears:
> 
>  [debug] upstream split a header line in FastCGI records
>  [error]  upstream sent too big header while reading response header
> from upstream

Only fastcgi_buffer_size is relevant here.
Could you set it to 512k and make request with debug log ?


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May  6 10:34:03 2011
From: nginx-forum at nginx.us (stalkercn)
Date: Fri, 06 May 2011 02:34:03 -0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <b3baffcc949cddf7dfb11f15dee198b7.NginxMailingListEnglish@forum.nginx.org>

http://www.toofiles.com/zh/rawfilesdownload-documents-log-nginx.html

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,192785,195989#msg-195989



From igor at sysoev.ru  Fri May  6 10:42:22 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 10:42:22 +0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <b3baffcc949cddf7dfb11f15dee198b7.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
	<b3baffcc949cddf7dfb11f15dee198b7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110506064222.GK85186@sysoev.ru>

On Fri, May 06, 2011 at 02:34:03AM -0400, stalkercn wrote:
> http://www.toofiles.com/zh/rawfilesdownload-documents-log-nginx.html

According to this log nginx uses 4k fastcgi_buffer_size.
Could you show your configuration ?


-- 
Igor Sysoev


From igor at sysoev.ru  Fri May  6 10:46:04 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 10:46:04 +0400
Subject: FastCGI server status
In-Reply-To: <BANLkTi=aJZWZ4FfTVe8kuszn3yni0ot75g@mail.gmail.com>
References: <BANLkTi=aJZWZ4FfTVe8kuszn3yni0ot75g@mail.gmail.com>
Message-ID: <20110506064604.GL85186@sysoev.ru>

On Thu, May 05, 2011 at 10:33:02PM -0300, Marcelo Cabral wrote:
> Hello,
> 
> I was wondering if I could monitor FastCGI responses (timers, success,
> failures, round robin stats, etc...) with nginx?
> Any hints on how I could get this info?

Only by logging variables:

$upstream_response_time
$upstream_status
$upstream_addr


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May  6 10:52:54 2011
From: nginx-forum at nginx.us (stalkercn)
Date: Fri, 06 May 2011 02:52:54 -0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <b3baffcc949cddf7dfb11f15dee198b7.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
	<b3baffcc949cddf7dfb11f15dee198b7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <352f1144fc8d4f12c32877f9005f9439.NginxMailingListEnglish@forum.nginx.org>

http://www.toofiles.com/zh/rawfilesdownload-documents-conf-nginx.html

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,192785,195996#msg-195996



From igor at sysoev.ru  Fri May  6 10:59:36 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 10:59:36 +0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <b3baffcc949cddf7dfb11f15dee198b7.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
	<b3baffcc949cddf7dfb11f15dee198b7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110506065936.GM85186@sysoev.ru>

On Fri, May 06, 2011 at 02:34:03AM -0400, stalkercn wrote:
> http://www.toofiles.com/zh/rawfilesdownload-documents-log-nginx.html

nginx does not merge locations as Apache does, so:

        location / {
            root   /home/www;
            index  index.php index.html;
        }

        location ~ \.php$ {
            root           /home/www;
            fastcgi_pass   127.0.0.1:9000;
            fastcgi_index  index.php;
            fastcgi_param SCRIPT_FILENAME /home/www$fastcgi_script_name;
            include        fastcgi_params;

            fastcgi_temp_file_write_size 10m;
            fastcgi_busy_buffers_size    512k;
            fastcgi_buffer_size          512k;
            fastcgi_buffers           16 512k;
            fastcgi_connect_timeout      300;
            fastcgi_send_timeout         300;
            fastcgi_read_timeout         300;
            fastcgi_intercept_errors     on;
            fastcgi_next_upstream error invalid_header timeout http_500;
        }


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May  6 11:22:23 2011
From: nginx-forum at nginx.us (stalkercn)
Date: Fri, 06 May 2011 03:22:23 -0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <352f1144fc8d4f12c32877f9005f9439.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
	<b3baffcc949cddf7dfb11f15dee198b7.NginxMailingListEnglish@forum.nginx.org>
	<352f1144fc8d4f12c32877f9005f9439.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <800b866ac3bd2851be3bd5dbc62e9c73.NginxMailingListEnglish@forum.nginx.org>

ooh a foolish fault! 
But as seem as proxy mode,between two contiguous lines in the debug
log,the time is 4-5 seconds,I dont know why?

eg:
2011/05/06 15:14:10 [debug] 86514#0: *101 http upstream request:
"/general/mytable/intel_view/index.php?"
2011/05/06 15:14:10 [debug] 86514#0: *101 http upstream dummy handler
2011/05/06 15:14:10 [debug] 86514#0: timer delta: 100
2011/05/06 15:14:10 [debug] 86514#0: posted events 0000000000000000
2011/05/06 15:14:10 [debug] 86514#0: worker cycle
2011/05/06 15:14:10 [debug] 86514#0: kevent timer: 299900, changes: 0
2011/05/06 15:14:14 [debug] 86514#0: kevent events: 1
2011/05/06 15:14:14 [debug] 86514#0: kevent: 8: ft:-1 fl:0020
ff:00000000 d:5816 ud:0000000801102150
2011/05/06 15:14:14 [debug] 86514#0: *101 http upstream request:
"/general/mytable/intel_view/index.php?"
2011/05/06 15:14:14 [debug] 86514#0: *101 http upstream process header
2011/05/06 15:14:14 [debug] 86514#0: *101 malloc:
0000000801302000:524288

the full log:
http://www.toofiles.com/zh/rawfilesdownload-documents-7z-nginx_log.html

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,192785,195999#msg-195999



From igor at sysoev.ru  Fri May  6 11:39:02 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 11:39:02 +0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <800b866ac3bd2851be3bd5dbc62e9c73.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
	<b3baffcc949cddf7dfb11f15dee198b7.NginxMailingListEnglish@forum.nginx.org>
	<352f1144fc8d4f12c32877f9005f9439.NginxMailingListEnglish@forum.nginx.org>
	<800b866ac3bd2851be3bd5dbc62e9c73.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110506073902.GA98651@sysoev.ru>

On Fri, May 06, 2011 at 03:22:23AM -0400, stalkercn wrote:
> ooh a foolish fault! 
> But as seem as proxy mode,between two contiguous lines in the debug
> log,the time is 4-5 seconds,I dont know why?
> 
> eg:
> 2011/05/06 15:14:10 [debug] 86514#0: *101 http upstream request:
> "/general/mytable/intel_view/index.php?"
> 2011/05/06 15:14:10 [debug] 86514#0: *101 http upstream dummy handler
> 2011/05/06 15:14:10 [debug] 86514#0: timer delta: 100
> 2011/05/06 15:14:10 [debug] 86514#0: posted events 0000000000000000
> 2011/05/06 15:14:10 [debug] 86514#0: worker cycle
> 2011/05/06 15:14:10 [debug] 86514#0: kevent timer: 299900, changes: 0
> 2011/05/06 15:14:14 [debug] 86514#0: kevent events: 1
> 2011/05/06 15:14:14 [debug] 86514#0: kevent: 8: ft:-1 fl:0020
> ff:00000000 d:5816 ud:0000000801102150
> 2011/05/06 15:14:14 [debug] 86514#0: *101 http upstream request:
> "/general/mytable/intel_view/index.php?"
> 2011/05/06 15:14:14 [debug] 86514#0: *101 http upstream process header
> 2011/05/06 15:14:14 [debug] 86514#0: *101 malloc:
> 0000000801302000:524288
> 
> the full log:
> http://www.toofiles.com/zh/rawfilesdownload-documents-7z-nginx_log.html

7z format is certainly not the best way to pack report logs.
There is old good gzip: it's enough.

As to delay, it seems that is is PHP delay.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May  6 12:02:45 2011
From: nginx-forum at nginx.us (stalkercn)
Date: Fri, 06 May 2011 04:02:45 -0400
Subject: How to avoid "too big header" issue?
In-Reply-To: <800b866ac3bd2851be3bd5dbc62e9c73.NginxMailingListEnglish@forum.nginx.org>
References: <12c7edea30d035b06c961d149e30db75.NginxMailingListEnglish@forum.nginx.org>
	<b3baffcc949cddf7dfb11f15dee198b7.NginxMailingListEnglish@forum.nginx.org>
	<352f1144fc8d4f12c32877f9005f9439.NginxMailingListEnglish@forum.nginx.org>
	<800b866ac3bd2851be3bd5dbc62e9c73.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <c72a7cd1d5adcdc82250e1d15a8dd52f.NginxMailingListEnglish@forum.nginx.org>

OK?I'll find out php problem,thanks anyway.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,192785,196007#msg-196007



From Richard.Kearsley at m247.com  Fri May  6 13:37:12 2011
From: Richard.Kearsley at m247.com (Richard Kearsley)
Date: Fri, 6 May 2011 09:37:12 +0000
Subject: aio info
In-Reply-To: <20110505071212.GA36559@sysoev.ru>
References: <EFC5A19DBF3A2E489904733B467D990B0EC5BE@office-server2.m247.local>
	<20110505071212.GA36559@sysoev.ru>
Message-ID: <EFC5A19DBF3A2E489904733B467D990B0EDC37@office-server2.m247.local>

Is aio used while reading files from proxy_cache?
I test aio on a file every 1s (it should be in ram buffers, so no disk io needed)
Usually it takes 1ms (or lower), but sometimes it spikes to over a second if disks are really busy
The max_aio_queue doesn't go above 3-4 during test
Any ideas the cause of spikes?


-----Original Message-----
From: Igor Sysoev [mailto:igor at sysoev.ru] 
Sent: 05 May 2011 08:12
To: nginx at nginx.org
Subject: Re: aio info

On Thu, May 05, 2011 at 05:05:37AM +0000, Richard Kearsley wrote:
> Hi
> A lot of talk about AIO recently - I actually didn't know it existed before, so I am trying it on freebsd8
> 
> # sysctl -a | grep aio
> kern.features.aio: 1
> vfs.aio.max_buf_aio: 16
> vfs.aio.max_aio_queue_per_proc: 256
> vfs.aio.max_aio_per_proc: 32
> vfs.aio.unloadable: 0
> vfs.aio.aiod_lifetime: 30000
> vfs.aio.aiod_timeout: 10000
> vfs.aio.num_buf_aio: 0
> vfs.aio.num_queue_count: 0
> vfs.aio.max_aio_queue: 1024
> vfs.aio.target_aio_procs: 4
> vfs.aio.num_aio_procs: 4
> vfs.aio.max_aio_procs: 32
> 
> Could anyone comment what these mean and how to go about tweaking for best performance (should any be increased?)

$ sysctl -d vfs.aio
vfs.aio: Async IO management
vfs.aio.max_buf_aio: Maximum buf aio requests per process (stored in the process)
vfs.aio.max_aio_queue_per_proc: Maximum queued aio requests per process (stored in the process)
vfs.aio.max_aio_per_proc: Maximum active aio requests per process (stored in the process)
vfs.aio.unloadable: Allow unload of aio (not recommended)
vfs.aio.aiod_lifetime: Maximum lifetime for idle aiod
vfs.aio.aiod_timeout: Timeout value for synchronous aio operations
vfs.aio.num_buf_aio: Number of aio requests presently handled by the buf subsystem
vfs.aio.num_queue_count: Number of queued aio requests
vfs.aio.max_aio_queue: Maximum number of aio requests to queue, globally
vfs.aio.target_aio_procs: Preferred number of ready kernel threads for async IO
vfs.aio.num_aio_procs: Number of presently active kernel threads for async IO
vfs.aio.max_aio_procs: Maximum number of kernel threads to use for handling async IO 

There are two types AIO operations inside FreeBSD kernel - raw file
AIO (buf aio) and usual file AIO that uses in-kernel threads/processes.
nginx uses only usual file AIO. Relevant sysctls are:

vfs.aio.max_aio_queue: 1024
vfs.aio.max_aio_queue_per_proc: 256
vfs.aio.max_aio_per_proc: 32
vfs.aio.num_queue_count: 0

vfs.aio.max_aio_procs: 32
vfs.aio.target_aio_procs: 4
vfs.aio.num_aio_procs: 4

vfs.aio.aiod_lifetime: 30000
vfs.aio.aiod_timeout: 10000

You may want to increase

vfs.aio.max_aio_procs
vfs.aio.max_aio_queue
vfs.aio.max_aio_queue_per_proc
vfs.aio.max_aio_per_proc

> Particularly.. vfs.aio.num_queue_count - does this monitor the amount of io waiting?

Yes.


-- 
Igor Sysoev

_______________________________________________
nginx mailing list
nginx at nginx.org
http://nginx.org/mailman/listinfo/nginx


From igor at sysoev.ru  Fri May  6 14:50:17 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 14:50:17 +0400
Subject: aio info
In-Reply-To: <EFC5A19DBF3A2E489904733B467D990B0EDC37@office-server2.m247.local>
References: <EFC5A19DBF3A2E489904733B467D990B0EC5BE@office-server2.m247.local>
	<20110505071212.GA36559@sysoev.ru>
	<EFC5A19DBF3A2E489904733B467D990B0EDC37@office-server2.m247.local>
Message-ID: <20110506105017.GD3100@sysoev.ru>

On Fri, May 06, 2011 at 09:37:12AM +0000, Richard Kearsley wrote:
> Is aio used while reading files from proxy_cache?

Yes, it works. For FreeBSD I recommend this settings:

    sendfile     on;
    aio          sendfile;
    tcp_nopush   on;
    read_ahead   512k;

> I test aio on a file every 1s (it should be in ram buffers, so no disk io needed)
> Usually it takes 1ms (or lower), but sometimes it spikes to over a second if disks are really busy
> The max_aio_queue doesn't go above 3-4 during test
> Any ideas the cause of spikes?

May be locks on VFS level. You can try this:

    open_file_cache   max=5000  inactive=20s;
    open_file_cache_valid       30s;
    open_file_cache_min_uses    3;


-- 
Igor Sysoev


From ionathan at gmail.com  Fri May  6 15:01:27 2011
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Fri, 6 May 2011 08:01:27 -0300
Subject: Virtualization
Message-ID: <BANLkTikHKd5PmbX6PzFduZ7ocizcWy3cgw@mail.gmail.com>

Hey!
I am trying to virtualize nginx in my own private cloud. Is there any
special consideration regard configuration that I have to consider?
Anyone did something similar? Regard throughput, anyone with some
numbers to share?

Thanks!

Jonathan


From Richard.Kearsley at m247.com  Fri May  6 16:52:51 2011
From: Richard.Kearsley at m247.com (Richard Kearsley)
Date: Fri, 6 May 2011 12:52:51 +0000
Subject: aio info
In-Reply-To: <20110506105017.GD3100@sysoev.ru>
References: <EFC5A19DBF3A2E489904733B467D990B0EC5BE@office-server2.m247.local>
	<20110505071212.GA36559@sysoev.ru>
	<EFC5A19DBF3A2E489904733B467D990B0EDC37@office-server2.m247.local>
	<20110506105017.GD3100@sysoev.ru>
Message-ID: <EFC5A19DBF3A2E489904733B467D990B0EDE0D@office-server2.m247.local>

Hi
Many thanks for your help, it is much appreciated.
I am still getting these pauses. They happen every 30 seconds on the dot.
What could be running/locking every 30 seconds? System disk (ad4) incurs usage at this point, and this should not be being used (ad5 and ad6 are the cache disks)
(note that I increased the valid/inactive times you exampled to 300s to rule that out)




From igor at sysoev.ru  Fri May  6 17:00:30 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 17:00:30 +0400
Subject: aio info
In-Reply-To: <EFC5A19DBF3A2E489904733B467D990B0EDE0D@office-server2.m247.local>
References: <EFC5A19DBF3A2E489904733B467D990B0EC5BE@office-server2.m247.local>
	<20110505071212.GA36559@sysoev.ru>
	<EFC5A19DBF3A2E489904733B467D990B0EDC37@office-server2.m247.local>
	<20110506105017.GD3100@sysoev.ru>
	<EFC5A19DBF3A2E489904733B467D990B0EDE0D@office-server2.m247.local>
Message-ID: <20110506130030.GE3100@sysoev.ru>

On Fri, May 06, 2011 at 12:52:51PM +0000, Richard Kearsley wrote:
> Hi
> Many thanks for your help, it is much appreciated.
> I am still getting these pauses. They happen every 30 seconds on the dot.
> What could be running/locking every 30 seconds? System disk (ad4) incurs usage at this point, and this should not be being used (ad5 and ad6 are the cache disks)
> (note that I increased the valid/inactive times you exampled to 300s to rule that out)

Could you run "top -S" and "systat -v 1" at time near to these peaks ?


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May  6 17:38:46 2011
From: nginx-forum at nginx.us (yellowbox9)
Date: Fri, 06 May 2011 09:38:46 -0400
Subject: empty reply from server
Message-ID: <4cd88446bdeea09b413fae442fa115e7.NginxMailingListEnglish@forum.nginx.org>

I'm setting up a new nginx environment and I'm having issues with 404's.
When I use curl to request a file that I know does not exist, nginx
responds back with an empty reply.... I would prefer a 404 not found
with the correct http status code.

My config:

user nobody;
worker_processes 4;
worker_rlimit_nofile 8192;
events
{
        worker_connections 4096;
}

http
{
        include mime.types;
        default_type text/plain;
        log_format nn '$host $remote_addr - $remote_user [$time_local] 
' '"$request" $status $bytes_sent ' '"$http_referer"
"$http_user_agent"';
        access_log /opt/nginx/logs/access.log nn;
        error_log /opt/nginx/logs/error.log;
        sendfile on;
        keepalive_timeout 0;
        server
        {
                listen 80 default;
                root /opt/nginx/html/;
                index index.html index.htm index.php index.php5;
                include /opt/nginx/conf/location_php.conf;
        }
        include /opt/nginx/conf/vhosts/*.conf;
}

The html root dir:

bash> ls /opt/nginx/html/
404.html  50x.html  empty.html  foobar  index.html  ping

The curl request:

curl -v http://localhost/index.html2
* About to connect() to localhost port 80
*   Trying 127.0.0.1... connected
* Connected to localhost (127.0.0.1) port 80
> GET /index.html2 HTTP/1.1
> User-Agent: curl/7.15.5 (i386-redhat-linux-gnu) libcurl/7.15.5
OpenSSL/0.9.8b zlib/1.2.3 libidn/0.6.5
> Host: localhost
> Accept: */*
> 
* Empty reply from server
* Connection #0 to host localhost left intact
curl: (52) Empty reply from server
* Closing connection #0

The access log:
localhost 127.0.0.1 - - [06/May/2011:09:35:17 -0400]  "GET /index.html2
HTTP/1.1" 200 0 "-" "curl/7.15.5 (i386-redhat-linux-gnu) libcurl/7.15.5
OpenSSL/0.9.8b zlib/1.2.3 libidn/0.6.5" - For some reason returning http
status code 200?

The error log:
2011/05/06 09:35:17 [error] 9879#0: *9312 open()
"/opt/nginx/html/index.html2" failed (2: No such file or directory),
client: 127.0.0.1, server: , request: "GET /index.html2 HTTP/1.1", host:
"localhost"

I don't understand what I'm doing wrong here... any thoughts?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196075,196075#msg-196075



From Richard.Kearsley at m247.com  Fri May  6 17:49:58 2011
From: Richard.Kearsley at m247.com (Richard Kearsley)
Date: Fri, 6 May 2011 13:49:58 +0000
Subject: aio info
In-Reply-To: <20110506130030.GE3100@sysoev.ru>
References: <EFC5A19DBF3A2E489904733B467D990B0EC5BE@office-server2.m247.local>
	<20110505071212.GA36559@sysoev.ru>
	<EFC5A19DBF3A2E489904733B467D990B0EDC37@office-server2.m247.local>
	<20110506105017.GD3100@sysoev.ru>
	<EFC5A19DBF3A2E489904733B467D990B0EDE0D@office-server2.m247.local>
	<20110506130030.GE3100@sysoev.ru>
Message-ID: <EFC5A19DBF3A2E489904733B467D990B0EDEC8@office-server2.m247.local>

Hi Igor,
Nothing stands out when looking at those stats (other than the disk %busy going from 50/60 straight up to 100). I.e. nothing is grabbing 100% cpu or something - there is barely 2% cpu usage and appears to drop during the 2-3 second lock period...
Note that during my test, nginx is rebuilding the proxy_cache index (1,000,000's of files that it has to read from disk to build index after a restart) - does something happen with the cache manager process every 30 seconds during rebuild?



From josh at joshparker.us  Fri May  6 18:33:52 2011
From: josh at joshparker.us (Parker, Joshua)
Date: Fri, 6 May 2011 10:33:52 -0400
Subject: Shared Hosting
Message-ID: <BANLkTikN3k2t_VU-AZ2Htb0Spc4rFq5TOQ@mail.gmail.com>

I have a dedicated server where i would like to offer shared hosting for
WordPress multisite installs. What would be the best configuration for the
virtual host files (more concerned with the server directive) since these
sites will need to utilize dedicated IP's especially if they want to use
domain mapping? Thanks for your help.

-- 
Joshua Parker
Church Management System - http://www.churchmember.us/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110506/5c2062b1/attachment.html>

From lists at ruby-forum.com  Fri May  6 19:22:15 2011
From: lists at ruby-forum.com (John Travolota)
Date: Fri, 06 May 2011 17:22:15 +0200
Subject: PHP virtual impossible on nginx?
Message-ID: <6cba0fd16dce811389b4bbcabf054e5b@ruby-forum.com>

It seems this is impossible on nginx,since php virtual is function
attached to apache,but still posting this in hope how someone managed to
make this work.For some reason a lot of scripts uses php virtual include
instead just php include and that means they cant work on nginx unless i
use SSI with SHTML extension.But i realy dont  like SHTML extension plus
i need to have support for SSI included on nginx configure for
that,would rather use php extension if possible.

-- 
Posted via http://www.ruby-forum.com/.


From nginx-forum at nginx.us  Fri May  6 20:16:55 2011
From: nginx-forum at nginx.us (jfix)
Date: Fri, 06 May 2011 12:16:55 -0400
Subject: trouble with forward proxying
Message-ID: <4e5c2db6e47b5d013dfbb23fa21a4b8a.NginxMailingListEnglish@forum.nginx.org>

Hello, I've looked far and wide but couldn't find an answer to my
problem.

Here's a quick description: an appserver must use the company's http
proxy for outgoing http requests, but cannot be configured. So I (tried)
set up nginx to act as a forward proxy, like this:

server {
	listen 9099;
	resolver 10.102.10.14;
	location / 		{
		proxy_set_header X-Real-IP $remote_addr;
		proxy_set_header Host $http_host;
		proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;

		proxy_pass
http://proxy.example.org:80$scheme://$http_host$request_uri;
	}
	
	error_log	/cygdrive/E/data/logs/forward-proxy-nginx-error.log	debug;
	access_log  /cygdrive/E/data/logs/forward-proxy-nginx-access.log 
combined buffer=32k;
}

I couldn't find any config examples where the parent proxy is used in
the proxy_pass directive. 

It seems like there is no other way to specify the company proxy than to
add it to the proxy_pass directive. 

The error message I get is this, apparently, nginx looks for a colon and
takes the rest as the port number:

2011/05/06 18:01:32 [error] 2420#0: *1 invalid port in upstream
"proxy.example.orghttp://www.google.com/", client: 127.0.0.1, server: ,
request: "GET http://www.google.com/ HTTP/1.1", host: "www.google.com"

If I change the proxy_pass directive to 

proxy_pass
http://proxy.example.org:80/$scheme://$http_host$request_uri;

it's the parent proxy that complains that a URL cannot start with a /
character:

The following error was encountered while trying to retrieve the URL: <a
href="/http://www.google.com/">/http://www.g
oogle.com/</a>

I'm at a loss how to continue. Could somebody please help?  Thanks very
much in advance, Jakob.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196115,196115#msg-196115



From lpr at dyb.fm  Fri May  6 20:20:24 2011
From: lpr at dyb.fm (Felipe Teofilo)
Date: Fri, 6 May 2011 13:20:24 -0300
Subject: 502 Bad Gateway
Message-ID: <978247EE-13A0-40BA-9886-61802681B105@dyb.fm>

Hello,


How can i change 502 bad gateway from all my vhosts, to do a unique erro page?


I want redirect to /var/html/www/errors/502/


I try put at vhost config, but some time it doesnt work.


Thanks,


Felipe aka lpr

From edhoprima at gmail.com  Fri May  6 20:31:06 2011
From: edhoprima at gmail.com (Edho P Arief)
Date: Fri, 6 May 2011 23:31:06 +0700
Subject: trouble with forward proxying
In-Reply-To: <4e5c2db6e47b5d013dfbb23fa21a4b8a.NginxMailingListEnglish@forum.nginx.org>
References: <4e5c2db6e47b5d013dfbb23fa21a4b8a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTi=RxuKnG9m=OK9pmSqfQYoe2AiyFQ@mail.gmail.com>

On Fri, May 6, 2011 at 11:16 PM, jfix <nginx-forum at nginx.us> wrote:
> I'm at a loss how to continue. Could somebody please help? ?Thanks very
> much in advance, Jakob.
>

I don't know, perhaps it's related with nginx's description?


From nginx-forum at nginx.us  Fri May  6 21:16:55 2011
From: nginx-forum at nginx.us (uid_b)
Date: Fri, 06 May 2011 13:16:55 -0400
Subject: Long URLs cause nginx to hang
Message-ID: <16d8c9361e848788e6d2e07ecc8596db.NginxMailingListEnglish@forum.nginx.org>

Hey everyone,

I'm having strange issues with _very_ long url's under Nginx 1.0.1 (also
tested under 0.8.54). Certain URL's with a large number of GET
parameters are failing. I wrote a quick test script which only prints 1
line (php), and added a massive GET parameter:

https://test/test.php?blah=aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa

This causes Nginx to hang indefinitely, the client receives absolutely
no data (verified by tcpdump). Though, if I cut the parameter in half,
it loads instantly.  Is there a configuration directive I need to tweak
in order to handle these massive URL's?

Thanks!

Ben

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196131,196131#msg-196131



From nginx-forum at nginx.us  Fri May  6 21:37:17 2011
From: nginx-forum at nginx.us (uid_b)
Date: Fri, 06 May 2011 13:37:17 -0400
Subject: Long URLs cause nginx to hang
In-Reply-To: <16d8c9361e848788e6d2e07ecc8596db.NginxMailingListEnglish@forum.nginx.org>
References: <16d8c9361e848788e6d2e07ecc8596db.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <d6c3b367ce228da86dd6e343ccfae7b8.NginxMailingListEnglish@forum.nginx.org>

update: to rule out fastcgi/php I ran the same test against a basic HTML
file with the same GET parameter, same problem.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196131,196138#msg-196138



From igor at sysoev.ru  Fri May  6 21:55:20 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 21:55:20 +0400
Subject: Long URLs cause nginx to hang
In-Reply-To: <d6c3b367ce228da86dd6e343ccfae7b8.NginxMailingListEnglish@forum.nginx.org>
References: <16d8c9361e848788e6d2e07ecc8596db.NginxMailingListEnglish@forum.nginx.org>
	<d6c3b367ce228da86dd6e343ccfae7b8.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110506175520.GB24743@sysoev.ru>

On Fri, May 06, 2011 at 01:37:17PM -0400, uid_b wrote:
> update: to rule out fastcgi/php I ran the same test against a basic HTML
> file with the same GET parameter, same problem.

Could you create a debug log of the HTML files request ?
http://nginx.org/en/docs/debugging_log.html


-- 
Igor Sysoev


From igor at sysoev.ru  Fri May  6 22:21:44 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 22:21:44 +0400
Subject: trouble with forward proxying
In-Reply-To: <4e5c2db6e47b5d013dfbb23fa21a4b8a.NginxMailingListEnglish@forum.nginx.org>
References: <4e5c2db6e47b5d013dfbb23fa21a4b8a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110506182144.GE24743@sysoev.ru>

On Fri, May 06, 2011 at 12:16:55PM -0400, jfix wrote:
> Hello, I've looked far and wide but couldn't find an answer to my
> problem.
> 
> Here's a quick description: an appserver must use the company's http
> proxy for outgoing http requests, but cannot be configured. So I (tried)
> set up nginx to act as a forward proxy, like this:
> 
> server {
> 	listen 9099;
> 	resolver 10.102.10.14;
> 	location / 		{
> 		proxy_set_header X-Real-IP $remote_addr;
> 		proxy_set_header Host $http_host;
> 		proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
> 
> 		proxy_pass
> http://proxy.example.org:80$scheme://$http_host$request_uri;
> 	}
> 	
> 	error_log	/cygdrive/E/data/logs/forward-proxy-nginx-error.log	debug;
> 	access_log  /cygdrive/E/data/logs/forward-proxy-nginx-access.log 
> combined buffer=32k;
> }
> 
> I couldn't find any config examples where the parent proxy is used in
> the proxy_pass directive. 
> 
> It seems like there is no other way to specify the company proxy than to
> add it to the proxy_pass directive. 
> 
> The error message I get is this, apparently, nginx looks for a colon and
> takes the rest as the port number:
> 
> 2011/05/06 18:01:32 [error] 2420#0: *1 invalid port in upstream
> "proxy.example.orghttp://www.google.com/", client: 127.0.0.1, server: ,
> request: "GET http://www.google.com/ HTTP/1.1", host: "www.google.com"
> 
> If I change the proxy_pass directive to 
> 
> proxy_pass
> http://proxy.example.org:80/$scheme://$http_host$request_uri;
> 
> it's the parent proxy that complains that a URL cannot start with a /
> character:
> 
> The following error was encountered while trying to retrieve the URL: <a
> href="/http://www.google.com/">/http://www.g
> oogle.com/</a>
> 
> I'm at a loss how to continue. Could somebody please help?  Thanks very
> much in advance, Jakob.

nginx is not intended to run as a forward proxy.


-- 
Igor Sysoev


From igor at sysoev.ru  Fri May  6 22:30:36 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 22:30:36 +0400
Subject: 502 Bad Gateway
In-Reply-To: <978247EE-13A0-40BA-9886-61802681B105@dyb.fm>
References: <978247EE-13A0-40BA-9886-61802681B105@dyb.fm>
Message-ID: <20110506183036.GA30428@sysoev.ru>

On Fri, May 06, 2011 at 01:20:24PM -0300, Felipe Teofilo wrote:
> Hello,
> 
> 
> How can i change 502 bad gateway from all my vhosts, to do a unique erro page?
> 
> 
> I want redirect to /var/html/www/errors/502/
> 
> 
> I try put at vhost config, but some time it doesnt work.

        error_page   500 502 503 504  /50x.html;
        location = /50x.html {
            root   /var/html/www/errors/502/;
        }


-- 
Igor Sysoev


From Brian.Akins at turner.com  Fri May  6 22:46:49 2011
From: Brian.Akins at turner.com (Akins, Brian)
Date: Fri, 06 May 2011 14:46:49 -0400
Subject: Lua pcre inside nginx
In-Reply-To: <4DC21B59.4010309@gmail.com>
Message-ID: <C9E9BA59.DD236%Brian.Akins@turner.com>

On 5/4/11 11:36 PM, "Chaos Wang" <chaoslawful at gmail.com> wrote:

> ngx_lua has fixed this problem, and can use pcre related extensions
> normally after v0.1.6rc5.

Thanks. Seems to work now.

-- 
Brian Akins



From nginx-forum at nginx.us  Fri May  6 22:52:43 2011
From: nginx-forum at nginx.us (TECK)
Date: Fri, 06 May 2011 14:52:43 -0400
Subject: MySQL and Nginx mystery
Message-ID: <371f900423de05e1a7d49f9888dcfe64.NginxMailingListEnglish@forum.nginx.org>

Hi all,

I have set a fastcgi upstream into Nginx configuration:

	upstream fastcgi {
		ip_hash;
		server			192.168.0.2:9000;
		server			192.168.0.3:9000;
		server			192.168.0.4:9000;
	}

Everything works perfect, except when I try to connect to MySQL, using
PHP.


$server = '192.168.0.40';
$username = 'tester';
$password = 'password';
mysql_connect($server, $username, $password);

I get a connection error telling me that I cannot connect to MySQL using
192.168.0.2!
(or whatever server I'm on it at that specific moment)

Current setup:
3 web servers, 2 set with heartbeat, nginx as start service
2 db servers, set with heartbeat and DRBD, mysqld as start service

Is there a header that I need to set into Nginx or this is a network
configuration issue that I need to alert the host guys about?
I did set the bind-address in my.cnf to 192.168.0.40 ip.

Thanks for your help.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196165,196165#msg-196165



From nginx-forum at nginx.us  Fri May  6 22:56:10 2011
From: nginx-forum at nginx.us (uid_b)
Date: Fri, 06 May 2011 14:56:10 -0400
Subject: Long URLs cause nginx to hang
In-Reply-To: <d6c3b367ce228da86dd6e343ccfae7b8.NginxMailingListEnglish@forum.nginx.org>
References: <16d8c9361e848788e6d2e07ecc8596db.NginxMailingListEnglish@forum.nginx.org>
	<d6c3b367ce228da86dd6e343ccfae7b8.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <9383ec9105c69a6fb02fbfbd183e9af9.NginxMailingListEnglish@forum.nginx.org>

I get 2 different outputs, depending on whether the connection is SSL or
not

Non-SSL:
2011/05/06 12:52:44 [debug] 13267#0: *48 accept: 192.168.10.21 fd:3
2011/05/06 12:52:44 [debug] 13267#0: *48 event timer add: 3:
120000:1304708084457
2011/05/06 12:52:44 [debug] 13267#0: *48 epoll add event: fd:3 op:1
ev:80000001

SSL:
2011/05/06 12:48:49 [debug] 12804#0: *16 accept: 192.168.10.21 fd:3
2011/05/06 12:48:49 [debug] 12804#0: *16 event timer add: 3:
120000:1304707849612
2011/05/06 12:48:49 [debug] 12804#0: *16 epoll add event: fd:3 op:1
ev:80000001
2011/05/06 12:48:49 [debug] 12804#0: *16 malloc: 0000000001440680:1296
2011/05/06 12:48:49 [debug] 12804#0: *16 posix_memalign:
000000000138BEB0:256 @16
2011/05/06 12:48:49 [debug] 12804#0: *16 malloc: 0000000001229710:1024
2011/05/06 12:48:49 [debug] 12804#0: *16 posix_memalign:
000000000151E9E0:4096 @16
2011/05/06 12:48:49 [debug] 12804#0: *16 http check ssl handshake
2011/05/06 12:48:49 [debug] 12804#0: *16 https ssl handshake: 0x16
2011/05/06 12:48:49 [debug] 12804#0: *16 SSL server name: "seismo"
2011/05/06 12:48:49 [debug] 12804#0: *16 SSL_do_handshake: -1
2011/05/06 12:48:49 [debug] 12804#0: *16 SSL_get_error: 2
2011/05/06 12:48:49 [debug] 12804#0: *16 SSL handshake handler: 0
2011/05/06 12:48:49 [debug] 12804#0: *16 SSL_do_handshake: 1
2011/05/06 12:48:49 [debug] 12804#0: *16 SSL: TLSv1, cipher:
"DHE-RSA-AES256-SHA SSLv3 Kx=DH Au=RSA Enc=AES(256) Mac=SHA1"
2011/05/06 12:48:49 [debug] 12804#0: *16 http process request line
2011/05/06 12:48:49 [debug] 12804#0: *16 SSL_read: -1
2011/05/06 12:48:49 [debug] 12804#0: *16 SSL_get_error: 2

If I remove the massive GET parameter:
2011/05/06 12:54:09 [debug] 13267#0: *69 accept: 192.168.10.21 fd:17
2011/05/06 12:54:09 [debug] 13267#0: *69 event timer add: 17:
120000:1304708169332
2011/05/06 12:54:09 [debug] 13267#0: *69 epoll add event: fd:17 op:1
ev:80000001
2011/05/06 12:54:09 [debug] 13267#0: *69 malloc: 00000000019EEE70:1296
2011/05/06 12:54:09 [debug] 13267#0: *69 posix_memalign:
0000000001B9B460:256 @16
2011/05/06 12:54:09 [debug] 13267#0: *69 malloc: 00000000019E0710:1024
2011/05/06 12:54:09 [debug] 13267#0: *69 posix_memalign:
0000000001C0AA40:4096 @16
2011/05/06 12:54:09 [debug] 13267#0: *69 http process request line
2011/05/06 12:54:09 [debug] 13267#0: *69 recv: fd:17 645 of 1024
2011/05/06 12:54:09 [debug] 13267#0: *69 http request line: "GET
/test.html HTTP/1.1"
2011/05/06 12:54:09 [debug] 13267#0: *69 http uri: "/test.html"
2011/05/06 12:54:09 [debug] 13267#0: *69 http args: ""
2011/05/06 12:54:09 [debug] 13267#0: *69 http exten: "html"
2011/05/06 12:54:09 [debug] 13267#0: *69 http process request header
line
2011/05/06 12:54:09 [debug] 13267#0: *69 http header: "Host: seismo"
2011/05/06 12:54:09 [debug] 13267#0: *69 http header: "User-Agent:
Mozilla/5.0 (X11; Linux x86_64; rv:2.0.1) Gecko/20100101 Firefox/4.0.1"
2011/05/06 12:54:09 [debug] 13267#0: *69 http header: "Accept:
text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8"
2011/05/06 12:54:09 [debug] 13267#0: *69 http header: "Accept-Language:
en-us,en;q=0.5"
2011/05/06 12:54:09 [debug] 13267#0: *69 http header: "Accept-Encoding:
gzip, deflate"
2011/05/06 12:54:09 [debug] 13267#0: *69 http header: "Accept-Charset:
ISO-8859-1,utf-8;q=0.7,*;q=0.7"
2011/05/06 12:54:09 [debug] 13267#0: *69 http header: "Keep-Alive: 115"
2011/05/06 12:54:09 [debug] 13267#0: *69 http header: "Connection:
keep-alive"
2011/05/06 12:54:09 [debug] 13267#0: *69 http header: "Cookie:
PHPSESSID=l3a89kpj6j0es6qbkkf1nd1p80;
__utma=16576436.82209771.1304707931.1304707931.1304707931.1;
__utmb=16576436.2.10.1304707931; __utmc=16576436;
__utmz=16576436.1304707931.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none)"
2011/05/06 12:54:09 [debug] 13267#0: *69 http header:
"If-Modified-Since: Fri, 06 May 2011 18:36:06 GMT"
2011/05/06 12:54:09 [debug] 13267#0: *69 http header done
2011/05/06 12:54:09 [debug] 13267#0: *69 event timer del: 17:
1304708169332
2011/05/06 12:54:09 [debug] 13267#0: *69 generic phase: 0
2011/05/06 12:54:09 [debug] 13267#0: *69 rewrite phase: 1
2011/05/06 12:54:09 [debug] 13267#0: *69 test location: "/"
2011/05/06 12:54:09 [debug] 13267#0: *69 test location: ~ "\.php"
2011/05/06 12:54:09 [debug] 13267#0: *69 test location: ~ "/\.ht"
2011/05/06 12:54:09 [debug] 13267#0: *69 using configuration "/sparkle"
2011/05/06 12:54:09 [debug] 13267#0: *69 http cl:-1 max:10485760
2011/05/06 12:54:09 [debug] 13267#0: *69 rewrite phase: 3
2011/05/06 12:54:09 [debug] 13267#0: *69 post rewrite phase: 4
2011/05/06 12:54:09 [debug] 13267#0: *69 generic phase: 5
2011/05/06 12:54:09 [debug] 13267#0: *69 generic phase: 6
2011/05/06 12:54:09 [debug] 13267#0: *69 generic phase: 7
2011/05/06 12:54:09 [debug] 13267#0: *69 access phase: 8
2011/05/06 12:54:09 [debug] 13267#0: *69 access phase: 9
2011/05/06 12:54:09 [debug] 13267#0: *69 post access phase: 10
2011/05/06 12:54:09 [debug] 13267#0: *69 try files phase: 11
2011/05/06 12:54:09 [debug] 13267#0: *69 http script var: "/test.html"
2011/05/06 12:54:09 [debug] 13267#0: *69 try to use file: "/test.html"
"/var/www/test.html"
2011/05/06 12:54:09 [debug] 13267#0: *69 try file uri: "/test.html"
2011/05/06 12:54:09 [debug] 13267#0: *69 content phase: 12
2011/05/06 12:54:09 [debug] 13267#0: *69 content phase: 13
2011/05/06 12:54:09 [debug] 13267#0: *69 content phase: 14
2011/05/06 12:54:09 [debug] 13267#0: *69 content phase: 15
2011/05/06 12:54:09 [debug] 13267#0: *69 content phase: 16
2011/05/06 12:54:09 [debug] 13267#0: *69 http filename:
"/var/www/test.html"
2011/05/06 12:54:09 [debug] 13267#0: *69 add cleanup: 0000000001C0B638
2011/05/06 12:54:09 [debug] 13267#0: *69 http static fd: 18
2011/05/06 12:54:09 [debug] 13267#0: *69 http set discard body
2011/05/06 12:54:09 [debug] 13267#0: *69 http ims:1304706966
lm:1304706966
2011/05/06 12:54:09 [debug] 13267#0: *69 HTTP/1.1 304 Not Modified
Server: nginx
Date: Fri, 06 May 2011 18:54:09 GMT
Last-Modified: Fri, 06 May 2011 18:36:06 GMT
Connection: keep-alive

2011/05/06 12:54:09 [debug] 13267#0: *69 write new buf t:1 f:0
0000000001C0B7D0, pos 0000000001C0B7D0, size: 151 file: 0, size: 0
2011/05/06 12:54:09 [debug] 13267#0: *69 http write filter: l:1 f:0
s:151
2011/05/06 12:54:09 [debug] 13267#0: *69 http write filter limit 0
2011/05/06 12:54:09 [debug] 13267#0: *69 writev: 151
2011/05/06 12:54:09 [debug] 13267#0: *69 http write filter
0000000000000000
2011/05/06 12:54:09 [debug] 13267#0: *69 http finalize request: 0,
"/sparkle/test.html?" a:1, c:1
2011/05/06 12:54:09 [debug] 13267#0: *69 set http keepalive handler
2011/05/06 12:54:09 [debug] 13267#0: *69 http close request
2011/05/06 12:54:09 [debug] 13267#0: *69 http log handler
2011/05/06 12:54:09 [debug] 13267#0: *69 run cleanup: 0000000001C0B638
2011/05/06 12:54:09 [debug] 13267#0: *69 file cleanup: fd:18
2011/05/06 12:54:09 [debug] 13267#0: *69 free: 0000000001C0AA40, unused:
273
2011/05/06 12:54:09 [debug] 13267#0: *69 event timer add: 17:
2000:1304708051332
2011/05/06 12:54:09 [debug] 13267#0: *69 free: 00000000019EEE70
2011/05/06 12:54:09 [debug] 13267#0: *69 free: 00000000019E0710
2011/05/06 12:54:09 [debug] 13267#0: *69 hc free: 0000000000000000 0
2011/05/06 12:54:09 [debug] 13267#0: *69 hc busy: 0000000000000000 0
2011/05/06 12:54:09 [debug] 13267#0: *69 tcp_nodelay
2011/05/06 12:54:09 [debug] 13267#0: *69 reusable connection: 1
2011/05/06 12:54:09 [debug] 13267#0: *69 post event 00000000019EE768
2011/05/06 12:54:09 [debug] 13267#0: *69 delete posted event
00000000019EE768
2011/05/06 12:54:09 [debug] 13267#0: *69 http keepalive handler
2011/05/06 12:54:09 [debug] 13267#0: *69 malloc: 00000000019E0710:1024
2011/05/06 12:54:09 [debug] 13267#0: *69 recv: fd:17 -1 of 1024
2011/05/06 12:54:09 [debug] 13267#0: *69 recv() not ready (11: Resource
temporarily unavailable)
2011/05/06 12:54:11 [debug] 13267#0: *69 event timer del: 17:
1304708051332
2011/05/06 12:54:11 [debug] 13267#0: *69 http keepalive handler
2011/05/06 12:54:11 [debug] 13267#0: *69 close http connection: 17
2011/05/06 12:54:11 [debug] 13267#0: *69 reusable connection: 0
2011/05/06 12:54:11 [debug] 13267#0: *69 free: 00000000019E0710
2011/05/06 12:54:11 [debug] 13267#0: *69 free: 0000000000000000
2011/05/06 12:54:11 [debug] 13267#0: *69 free: 0000000001B9AC50, unused:
8
2011/05/06 12:54:11 [debug] 13267#0: *69 free: 0000000001B9B460, unused:
128

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196131,196167#msg-196167



From igor at sysoev.ru  Fri May  6 23:00:10 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 23:00:10 +0400
Subject: Long URLs cause nginx to hang
In-Reply-To: <9383ec9105c69a6fb02fbfbd183e9af9.NginxMailingListEnglish@forum.nginx.org>
References: <16d8c9361e848788e6d2e07ecc8596db.NginxMailingListEnglish@forum.nginx.org>
	<d6c3b367ce228da86dd6e343ccfae7b8.NginxMailingListEnglish@forum.nginx.org>
	<9383ec9105c69a6fb02fbfbd183e9af9.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110506190009.GC30428@sysoev.ru>

On Fri, May 06, 2011 at 02:56:10PM -0400, uid_b wrote:
> I get 2 different outputs, depending on whether the connection is SSL or
> not

> 2011/05/06 12:54:09 [debug] 13267#0: *69 http header: "User-Agent:
> Mozilla/5.0 (X11; Linux x86_64; rv:2.0.1) Gecko/20100101 Firefox/4.0.1"
> 2011/05/06 12:54:09 [debug] 13267#0: *69 http header: "Accept:
> text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8"

Did you try other browser ?


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May  6 23:14:22 2011
From: nginx-forum at nginx.us (uid_b)
Date: Fri, 06 May 2011 15:14:22 -0400
Subject: Long URLs cause nginx to hang
In-Reply-To: <20110506190009.GC30428@sysoev.ru>
References: <20110506190009.GC30428@sysoev.ru>
Message-ID: <f76f6390cd6ffe0a7f3b380f2fa73a94.NginxMailingListEnglish@forum.nginx.org>

Igor Sysoev Wrote:
-------------------------------------------------------
> Did you try other browser ?

Yes, same message from both Firefox 4.0.1 and Chrome 11.0

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196131,196171#msg-196171



From igor at sysoev.ru  Fri May  6 23:27:28 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 23:27:28 +0400
Subject: Long URLs cause nginx to hang
In-Reply-To: <9383ec9105c69a6fb02fbfbd183e9af9.NginxMailingListEnglish@forum.nginx.org>
References: <16d8c9361e848788e6d2e07ecc8596db.NginxMailingListEnglish@forum.nginx.org>
	<d6c3b367ce228da86dd6e343ccfae7b8.NginxMailingListEnglish@forum.nginx.org>
	<9383ec9105c69a6fb02fbfbd183e9af9.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110506192728.GD30428@sysoev.ru>

On Fri, May 06, 2011 at 02:56:10PM -0400, uid_b wrote:
> I get 2 different outputs, depending on whether the connection is SSL or
> not
> 
> Non-SSL:
> 2011/05/06 12:52:44 [debug] 13267#0: *48 accept: 192.168.10.21 fd:3
> 2011/05/06 12:52:44 [debug] 13267#0: *48 event timer add: 3:
> 120000:1304708084457
> 2011/05/06 12:52:44 [debug] 13267#0: *48 epoll add event: fd:3 op:1
> ev:80000001

nginx does not receive any data and it should close connection after 120s.
Did you run tcpdump on server side ?


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May  6 23:39:50 2011
From: nginx-forum at nginx.us (uid_b)
Date: Fri, 06 May 2011 15:39:50 -0400
Subject: Long URLs cause nginx to hang
In-Reply-To: <f76f6390cd6ffe0a7f3b380f2fa73a94.NginxMailingListEnglish@forum.nginx.org>
References: <20110506190009.GC30428@sysoev.ru>
	<f76f6390cd6ffe0a7f3b380f2fa73a94.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <ed39c00150ded78934110d0c511c37e5.NginxMailingListEnglish@forum.nginx.org>

This appears to be an issue with our VPN, I have no idea why everything
works except for massive URL's but I don't think it's related to nginx
in any way.

Thanks for your help!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196131,196177#msg-196177



From nginx-forum at nginx.us  Fri May  6 23:50:08 2011
From: nginx-forum at nginx.us (TECK)
Date: Fri, 06 May 2011 15:50:08 -0400
Subject: MySQL and Nginx mystery
In-Reply-To: <371f900423de05e1a7d49f9888dcfe64.NginxMailingListEnglish@forum.nginx.org>
References: <371f900423de05e1a7d49f9888dcfe64.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <1644bcfa3d142f1496577fc567ace985.NginxMailingListEnglish@forum.nginx.org>

What I actually try to do is: force Nginx to show one IP to the MySQL
server, no matter on what server the user is.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196165,196181#msg-196181



From igor at sysoev.ru  Fri May  6 23:51:58 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 6 May 2011 23:51:58 +0400
Subject: Long URLs cause nginx to hang
In-Reply-To: <ed39c00150ded78934110d0c511c37e5.NginxMailingListEnglish@forum.nginx.org>
References: <20110506190009.GC30428@sysoev.ru>
	<f76f6390cd6ffe0a7f3b380f2fa73a94.NginxMailingListEnglish@forum.nginx.org>
	<ed39c00150ded78934110d0c511c37e5.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110506195158.GA40054@sysoev.ru>

On Fri, May 06, 2011 at 03:39:50PM -0400, uid_b wrote:
> This appears to be an issue with our VPN, I have no idea why everything
> works except for massive URL's but I don't think it's related to nginx
> in any way.

Try to search for "path mtu discovery vpn".


-- 
Igor Sysoev


From francis at daoine.org  Sat May  7 01:06:26 2011
From: francis at daoine.org (Francis Daly)
Date: Fri, 6 May 2011 22:06:26 +0100
Subject: MySQL and Nginx mystery
In-Reply-To: <1644bcfa3d142f1496577fc567ace985.NginxMailingListEnglish@forum.nginx.org>
References: <371f900423de05e1a7d49f9888dcfe64.NginxMailingListEnglish@forum.nginx.org>
	<1644bcfa3d142f1496577fc567ace985.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110506210626.GD12564@craic.sysops.org>

On Fri, May 06, 2011 at 03:50:08PM -0400, TECK wrote:

Hi there,

> What I actually try to do is: force Nginx to show one IP to the MySQL
> server, no matter on what server the user is.

In your setup, nginx isn't talking to mysql at all.

It's the fastcgi server that runs the php code that connects to mysql. So
you'll want to look there for the resolution.

Good luck,

	f
-- 
Francis Daly        francis at daoine.org


From nginx-forum at nginx.us  Sat May  7 16:49:43 2011
From: nginx-forum at nginx.us (bluescreen303)
Date: Sat, 07 May 2011 08:49:43 -0400
Subject: proxy caching
Message-ID: <4c8ed01f5f8401ff4d3e3634bc42488a.NginxMailingListEnglish@forum.nginx.org>

Hi all,

I would like to use nginx to cache some responses from my backends.
There are basically 2 types of responses I would like to cache.

- public images that never expire
- private content

The first one was easy, got that working already. Since my application
uses timestamps in the html when referring to images, changes to them
will be instantly picked up.

The second one is way harder.
Responses can become quite large (1 upto 10Mb).
My backend application is putting etags on the responses so it can
return 304s if nothing changed.
Also, it puts a 5min max-age cache-control on it,
Lastly, responses are marked private, because they are behind http basic
authentication (handled by the backend app).

What I would like nginx to do is the following:
It should store the first response for a certain url, using the current
http auth in the key (so it won't accidentally become publicly
available).
If a second request comes in for the same url with the same http auth
within 5 minutes: serve from cache.
If a second request comes in for the same url with the same http auth
_after_ 5 mintutes: ask backend using the if-not-modified header.
- if modified, store the new response.
- if 304 not modified, respond from cache, update cache to be valid for
5 minutes again.
If a request comes in for the same url but with other http auth: ignore
cache and ask backend.

The http auth stuff seems solvable by using the http auth in the
cache_key.
The other 2 things I would like to solve are:

- I can't find clear info about nginx's handling of etag and
if-not-modified. Will it work as described above out-of-the-box?
It's especially important that nginx uses if-not-modified when querying
the backend, even if the client request itself didn't contain it.
Ofcourse if the client request did include it, it should receive a nice
304 too.

- The backend sends Cache-Control: private for these responses. I would
like nginx to cache them though.
Ignoring the Cache-Control header doesn't seem to work (because without
cache-control at all, caching seems disabled)
Another option might be to have the backend set cache-control to public
and have nginx set it back to private, but that sounds a bit hacky.

Does anyone have any suggestions about how to proceed?
Thanks a lot!

Mathijs

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196280,196280#msg-196280



From mdounin at mdounin.ru  Sun May  8 02:34:23 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sun, 8 May 2011 02:34:23 +0400
Subject: empty reply from server
In-Reply-To: <4cd88446bdeea09b413fae442fa115e7.NginxMailingListEnglish@forum.nginx.org>
References: <4cd88446bdeea09b413fae442fa115e7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110507223423.GC42265@mdounin.ru>

Hello!

On Fri, May 06, 2011 at 09:38:46AM -0400, yellowbox9 wrote:

> I'm setting up a new nginx environment and I'm having issues with 404's.
> When I use curl to request a file that I know does not exist, nginx
> responds back with an empty reply.... I would prefer a 404 not found
> with the correct http status code.

You will be surprised, but nginx returns 404 in such cases by default.

> My config:
> 
> user nobody;
> worker_processes 4;
> worker_rlimit_nofile 8192;
> events
> {
>         worker_connections 4096;
> }
> 
> http
> {
>         include mime.types;
>         default_type text/plain;
>         log_format nn '$host $remote_addr - $remote_user [$time_local] 
> ' '"$request" $status $bytes_sent ' '"$http_referer"
> "$http_user_agent"';
>         access_log /opt/nginx/logs/access.log nn;
>         error_log /opt/nginx/logs/error.log;

It's good idea to specify error_log at *global* level, not http.  
When defined at http level it will be only used for http-related 
problems, but not for generic problems.

E.g. with your config information about abnormally terminated 
workers won't reach error_log you specified, but will go to 
compiled-in default error log instead.

>         sendfile on;
>         keepalive_timeout 0;
>         server
>         {
>                 listen 80 default;
>                 root /opt/nginx/html/;
>                 index index.html index.htm index.php index.php5;
>                 include /opt/nginx/conf/location_php.conf;

Please show contents of the included file.

>         }
>         include /opt/nginx/conf/vhosts/*.conf;

And please show contents of these includes, if there are any.

> }
> 
> The html root dir:
> 
> bash> ls /opt/nginx/html/
> 404.html  50x.html  empty.html  foobar  index.html  ping
> 
> The curl request:
> 
> curl -v http://localhost/index.html2
> * About to connect() to localhost port 80
> *   Trying 127.0.0.1... connected
> * Connected to localhost (127.0.0.1) port 80
> > GET /index.html2 HTTP/1.1
> > User-Agent: curl/7.15.5 (i386-redhat-linux-gnu) libcurl/7.15.5
> OpenSSL/0.9.8b zlib/1.2.3 libidn/0.6.5
> > Host: localhost
> > Accept: */*
> > 
> * Empty reply from server
> * Connection #0 to host localhost left intact
> curl: (52) Empty reply from server
> * Closing connection #0
> 
> The access log:
> localhost 127.0.0.1 - - [06/May/2011:09:35:17 -0400]  "GET /index.html2
> HTTP/1.1" 200 0 "-" "curl/7.15.5 (i386-redhat-linux-gnu) libcurl/7.15.5
> OpenSSL/0.9.8b zlib/1.2.3 libidn/0.6.5" - For some reason returning http
> status code 200?
> 
> The error log:
> 2011/05/06 09:35:17 [error] 9879#0: *9312 open()
> "/opt/nginx/html/index.html2" failed (2: No such file or directory),
> client: 127.0.0.1, server: , request: "GET /index.html2 HTTP/1.1", host:
> "localhost"
> 
> I don't understand what I'm doing wrong here... any thoughts?

Empty response indicate either "return 444" somewhere in your 
config or some error occured (and it's impossible to return error 
to the client).

Status 200 in your access log indicate that your set some 
error_page processing somewhere in your config.

It's impossible to say anything for sure from information you 
provided (basically there is no full error log, and no full 
config), but I assume you did something like

    error_page 404 =200 /error;

    location /error {
        return 444;
    }

which will lead to the exactly the same results as you described.

Maxim Dounin


From mdounin at mdounin.ru  Sun May  8 03:29:31 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sun, 8 May 2011 03:29:31 +0400
Subject: proxy caching
In-Reply-To: <4c8ed01f5f8401ff4d3e3634bc42488a.NginxMailingListEnglish@forum.nginx.org>
References: <4c8ed01f5f8401ff4d3e3634bc42488a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110507232930.GD42265@mdounin.ru>

Hello!

On Sat, May 07, 2011 at 08:49:43AM -0400, bluescreen303 wrote:

> Hi all,
> 
> I would like to use nginx to cache some responses from my backends.
> There are basically 2 types of responses I would like to cache.
> 
> - public images that never expire
> - private content
> 
> The first one was easy, got that working already. Since my application
> uses timestamps in the html when referring to images, changes to them
> will be instantly picked up.
> 
> The second one is way harder.
> Responses can become quite large (1 upto 10Mb).
> My backend application is putting etags on the responses so it can
> return 304s if nothing changed.
> Also, it puts a 5min max-age cache-control on it,
> Lastly, responses are marked private, because they are behind http basic
> authentication (handled by the backend app).
> 
> What I would like nginx to do is the following:
> It should store the first response for a certain url, using the current
> http auth in the key (so it won't accidentally become publicly
> available).
> If a second request comes in for the same url with the same http auth
> within 5 minutes: serve from cache.
> If a second request comes in for the same url with the same http auth
> _after_ 5 mintutes: ask backend using the if-not-modified header.
> - if modified, store the new response.
> - if 304 not modified, respond from cache, update cache to be valid for
> 5 minutes again.
> If a request comes in for the same url but with other http auth: ignore
> cache and ask backend.
> 
> The http auth stuff seems solvable by using the http auth in the
> cache_key.
> The other 2 things I would like to solve are:
> 
> - I can't find clear info about nginx's handling of etag and
> if-not-modified. Will it work as described above out-of-the-box?
> It's especially important that nginx uses if-not-modified when querying
> the backend, even if the client request itself didn't contain it.
> Ofcourse if the client request did include it, it should receive a nice
> 304 too.

Right now nginx can't use conditional requests to refresh cache.

> - The backend sends Cache-Control: private for these responses. I would
> like nginx to cache them though.
> Ignoring the Cache-Control header doesn't seem to work (because without
> cache-control at all, caching seems disabled)
> Another option might be to have the backend set cache-control to public
> and have nginx set it back to private, but that sounds a bit hacky.

Solution is to ignore Cache-Control header and enable caching 
explicitly with proxy_cache_valid directive.

Maxim Dounin


From luky-37 at hotmail.com  Sun May  8 20:17:11 2011
From: luky-37 at hotmail.com (Lukas Tribus)
Date: Sun, 8 May 2011 18:17:11 +0200
Subject: proxy caching
In-Reply-To: <20110507232930.GD42265@mdounin.ru>
References: <4c8ed01f5f8401ff4d3e3634bc42488a.NginxMailingListEnglish@forum.nginx.org>,
	<20110507232930.GD42265@mdounin.ru>
Message-ID: <DUB107-w65D4A9F3C2248C27D21531ED850@phx.gbl>


Hi,

give varnish [1] a look, its probably better suited for such a scenario.


[1] http://www.varnish-cache.org/



> Date: Sun, 8 May 2011 03:29:31 +0400
> From: mdounin at mdounin.ru
> To: nginx at nginx.org
> Subject: Re: proxy caching
> 
> Hello!
> 
> On Sat, May 07, 2011 at 08:49:43AM -0400, bluescreen303 wrote:
> 
> > Hi all,
> > 
> > I would like to use nginx to cache some responses from my backends.
> > There are basically 2 types of responses I would like to cache.
> > 
> > - public images that never expire
> > - private content
> > 
> > The first one was easy, got that working already. Since my application
> > uses timestamps in the html when referring to images, changes to them
> > will be instantly picked up.
> > 
> > The second one is way harder.
> > Responses can become quite large (1 upto 10Mb).
> > My backend application is putting etags on the responses so it can
> > return 304s if nothing changed.
> > Also, it puts a 5min max-age cache-control on it,
> > Lastly, responses are marked private, because they are behind http basic
> > authentication (handled by the backend app).
> > 
> > What I would like nginx to do is the following:
> > It should store the first response for a certain url, using the current
> > http auth in the key (so it won't accidentally become publicly
> > available).
> > If a second request comes in for the same url with the same http auth
> > within 5 minutes: serve from cache.
> > If a second request comes in for the same url with the same http auth
> > _after_ 5 mintutes: ask backend using the if-not-modified header.
> > - if modified, store the new response.
> > - if 304 not modified, respond from cache, update cache to be valid for
> > 5 minutes again.
> > If a request comes in for the same url but with other http auth: ignore
> > cache and ask backend.
> > 
> > The http auth stuff seems solvable by using the http auth in the
> > cache_key.
> > The other 2 things I would like to solve are:
> > 
> > - I can't find clear info about nginx's handling of etag and
> > if-not-modified. Will it work as described above out-of-the-box?
> > It's especially important that nginx uses if-not-modified when querying
> > the backend, even if the client request itself didn't contain it.
> > Ofcourse if the client request did include it, it should receive a nice
> > 304 too.
> 
> Right now nginx can't use conditional requests to refresh cache.
> 
> > - The backend sends Cache-Control: private for these responses. I would
> > like nginx to cache them though.
> > Ignoring the Cache-Control header doesn't seem to work (because without
> > cache-control at all, caching seems disabled)
> > Another option might be to have the backend set cache-control to public
> > and have nginx set it back to private, but that sounds a bit hacky.
> 
> Solution is to ignore Cache-Control header and enable caching 
> explicitly with proxy_cache_valid directive.
> 
> Maxim Dounin
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110508/606a977a/attachment.html>

From nginx-forum at nginx.us  Sun May  8 23:15:28 2011
From: nginx-forum at nginx.us (zildjohn01)
Date: Sun, 08 May 2011 15:15:28 -0400
Subject: nginx silently drops response body of POST depending on its length
Message-ID: <221afc752b74444b2aa51a8b264f930c.NginxMailingListEnglish@forum.nginx.org>

I've come across an incredibly weird problem with POST responses using
nginx 1.0.1 and uWSGI 0.9.7.2 on CentOS 5.6.

uWSGI successfully sends the response body to nginx, but nginx seems to
ignore it if both (1) the request method is POST, and (2) the response
length is less than 4052 bytes. The issue doesn't occur when serving
directly using `uwsgi --http`, so I'm pretty sure the problem is
something to do with nginx.

To explain things in code, this Python results in the desired response:

....def app_uwsgi(environ, start_response):
........start_response('200 OK', [('Content-type', 'text/html')])
........return [b'x' * 4052]

while this results in zero bytes being sent to the client for POST
requests:

....def app_uwsgi(environ, start_response):
........start_response('200 OK', [('Content-type', 'text/html')])
........return [b'x' * 4051]

I stripped down to a minimal nginx.conf, and the issue still occurs:

....events {
........worker_connections  1024;
....}

....http {
........upstream uwsgi_connect {
............server  127.0.0.1:39275;
........}

........server {
............listen       20101;
............server_name  domain.net;

............location / {
................uwsgi_pass  uwsgi_connect;
................include     uwsgi_params;
............}
........}
....}

Does anyone have any clue what could be causing this behavior?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196581,196581#msg-196581



From nginx-forum at nginx.us  Mon May  9 07:12:29 2011
From: nginx-forum at nginx.us (bigplum)
Date: Sun, 08 May 2011 23:12:29 -0400
Subject: nginx proxying do not locally save/buffer uploaded files?
In-Reply-To: <20110504211549.GL32595@reaktio.net>
References: <20110504211549.GL32595@reaktio.net>
Message-ID: <e7b890f79df4c9ba63ebfac719f6b799.NginxMailingListEnglish@forum.nginx.org>

If we set proxy_store and proxy_cache to off, will nginx cache uploaded
files?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,195656,196629#msg-196629



From roberto at unbit.it  Mon May  9 09:41:42 2011
From: roberto at unbit.it (Roberto De Ioris)
Date: Mon, 9 May 2011 07:41:42 +0200 (CEST)
Subject: nginx silently drops response body of POST depending on its length
In-Reply-To: <221afc752b74444b2aa51a8b264f930c.NginxMailingListEnglish@forum.nginx.
	org>
References: <221afc752b74444b2aa51a8b264f930c.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <d729ad9250d81b3ba6e5d54e940a9644.squirrel@manage.unbit.it>


> I've come across an incredibly weird problem with POST responses using
> nginx 1.0.1 and uWSGI 0.9.7.2 on CentOS 5.6.
>
> uWSGI successfully sends the response body to nginx, but nginx seems to
> ignore it if both (1) the request method is POST, and (2) the response
> length is less than 4052 bytes. The issue doesn't occur when serving
> directly using `uwsgi --http`, so I'm pretty sure the problem is
> something to do with nginx.
>
> To explain things in code, this Python results in the desired response:
>
> ....def app_uwsgi(environ, start_response):
> ........start_response('200 OK', [('Content-type', 'text/html')])
> ........return [b'x' * 4052]
>
> while this results in zero bytes being sent to the client for POST
> requests:
>
> ....def app_uwsgi(environ, start_response):
> ........start_response('200 OK', [('Content-type', 'text/html')])
> ........return [b'x' * 4051]
>
> I stripped down to a minimal nginx.conf, and the issue still occurs:
>
> ....events {
> ........worker_connections  1024;
> ....}
>
> ....http {
> ........upstream uwsgi_connect {
> ............server  127.0.0.1:39275;
> ........}
>
> ........server {
> ............listen       20101;
> ............server_name  domain.net;
>
> ............location / {
> ................uwsgi_pass  uwsgi_connect;
> ................include     uwsgi_params;
> ............}
> ........}
> ....}
>
>


Be sure to read POST body in your WSGI app if Content-Length > 0,
otherwise your socket queue will get clobbered (nginx will receive a close
when there is still data available in the queue)

-- 
Roberto De Ioris
http://unbit.it


From jeffrpang at gmail.com  Mon May  9 10:18:09 2011
From: jeffrpang at gmail.com (Jeff Pang)
Date: Mon, 9 May 2011 14:18:09 +0800
Subject: server_name for https
Message-ID: <BANLkTim=GtK0AnJvb_XiTCt6nGmjsNpaTQ@mail.gmail.com>

Hello,

Since HTTPS doesn't support vhost, shall we have the need to put the
server_name command in the server section? for example,

         server {
                listen       443;
                server_name  xxx.example.com;
                ....
         }

Thanks.


-- 
Jeff Pang
www.DNSbed.com


From pasik at iki.fi  Mon May  9 10:34:07 2011
From: pasik at iki.fi (Pasi =?iso-8859-1?Q?K=E4rkk=E4inen?=)
Date: Mon, 9 May 2011 09:34:07 +0300
Subject: nginx proxying do not locally save/buffer uploaded files?
In-Reply-To: <e7b890f79df4c9ba63ebfac719f6b799.NginxMailingListEnglish@forum.nginx.org>
References: <20110504211549.GL32595@reaktio.net>
	<e7b890f79df4c9ba63ebfac719f6b799.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110509063407.GZ32595@reaktio.net>

On Sun, May 08, 2011 at 11:12:29PM -0400, bigplum wrote:
> If we set proxy_store and proxy_cache to off, will nginx cache uploaded
> files?
> 

proxy_cache only mentions "backend" .. so I think it's not for uploads..

-- Pasi



From cedric.jeanneret at camptocamp.com  Mon May  9 11:06:27 2011
From: cedric.jeanneret at camptocamp.com (Cedric Jeanneret)
Date: Mon, 9 May 2011 09:06:27 +0200
Subject: server_name for https
In-Reply-To: <BANLkTim=GtK0AnJvb_XiTCt6nGmjsNpaTQ@mail.gmail.com>
References: <BANLkTim=GtK0AnJvb_XiTCt6nGmjsNpaTQ@mail.gmail.com>
Message-ID: <20110509090627.68905d0d@cholatse.wrk.lsn.camptocamp.com>

On Mon, 9 May 2011 14:18:09 +0800
Jeff Pang <jeffrpang at gmail.com> wrote:

> Hello,
> 
> Since HTTPS doesn't support vhost, shall we have the need to put the
> server_name command in the server section? for example,
> 
>          server {
>                 listen       443;
>                 server_name  xxx.example.com;
>                 ....
>          }
> 
> Thanks.
> 
> 

o_O since when does HTTPS not support vhosts ?

the only problem is when having multiple SSL vhosts on the same IP - the certificate has to cover all those vhosts.

To answer your question, you shall provide:
server_name $_;

it's the default vhost

Cheers,

C.

-- 
C?dric Jeanneret                 |  System Administrator
021 619 10 32                    |  Camptocamp SA
cedric.jeanneret at camptocamp.com  |  PSE-A / EPFL
www.camptocamp.com               | www.github.com/camptocamp
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://nginx.org/pipermail/nginx/attachments/20110509/975ca98c/attachment.pgp>

From jeffrpang at gmail.com  Mon May  9 11:39:11 2011
From: jeffrpang at gmail.com (Jeff Pang)
Date: Mon, 9 May 2011 15:39:11 +0800
Subject: server_name for https
In-Reply-To: <20110509090627.68905d0d@cholatse.wrk.lsn.camptocamp.com>
References: <BANLkTim=GtK0AnJvb_XiTCt6nGmjsNpaTQ@mail.gmail.com>
	<20110509090627.68905d0d@cholatse.wrk.lsn.camptocamp.com>
Message-ID: <BANLkTimLYRBN5CdG5b5FDEJx9BM+Mymz6w@mail.gmail.com>

2011/5/9 Cedric Jeanneret <cedric.jeanneret at camptocamp.com>:
> On Mon, 9 May 2011 14:18:09 +0800
> Jeff Pang <jeffrpang at gmail.com> wrote:
>
>> Hello,
>>
>> Since HTTPS doesn't support vhost, shall we have the need to put the
>> server_name command in the server section? for example,
>>
>> ? ? ? ? ?server {
>> ? ? ? ? ? ? ? ? listen ? ? ? 443;
>> ? ? ? ? ? ? ? ? server_name ?xxx.example.com;
>> ? ? ? ? ? ? ? ? ....
>> ? ? ? ? ?}
>>
>> Thanks.
>>
>>
>
> o_O since when does HTTPS not support vhosts ?
>
> the only problem is when having multiple SSL vhosts on the same IP - the certificate has to cover all those vhosts.
>
> To answer your question, you shall provide:
> server_name $_;
>
> it's the default vhost
>


OK That's right thanks.
I have misunderstood the SSL Proxy maybe, now I got it.



-- 
Jeff Pang
www.DNSbed.com


From Richard.Kearsley at m247.com  Mon May  9 13:33:18 2011
From: Richard.Kearsley at m247.com (Richard Kearsley)
Date: Mon, 9 May 2011 09:33:18 +0000
Subject: aio info
In-Reply-To: <20110506130030.GE3100@sysoev.ru>
References: <EFC5A19DBF3A2E489904733B467D990B0EC5BE@office-server2.m247.local>
	<20110505071212.GA36559@sysoev.ru>
	<EFC5A19DBF3A2E489904733B467D990B0EDC37@office-server2.m247.local>
	<20110506105017.GD3100@sysoev.ru>
	<EFC5A19DBF3A2E489904733B467D990B0EDE0D@office-server2.m247.local>
	<20110506130030.GE3100@sysoev.ru>
Message-ID: <EFC5A19DBF3A2E489904733B467D990B0F4772@office-server2.m247.local>

Running nginx on freebsd 8.2 with 16 workers and aio on, when the server encounters heavy writes, all processes are in D state
Aio queue is way below system/process limit
A request for small file (said file will be in fs buffer/cache) every second varies wildly from 1ms to over 1s

>From man ps:
D       Marks a process in disk (or other short term, uninter-ruptible) wait.

What's causing the lock?

Many thanks
Richard


From pascal.nitsche at fansubcode.org  Mon May  9 13:48:55 2011
From: pascal.nitsche at fansubcode.org (Pascal Nitsche)
Date: Mon, 09 May 2011 11:48:55 +0200
Subject: server_name for https
In-Reply-To: <20110509090627.68905d0d@cholatse.wrk.lsn.camptocamp.com>
References: <BANLkTim=GtK0AnJvb_XiTCt6nGmjsNpaTQ@mail.gmail.com>
	<20110509090627.68905d0d@cholatse.wrk.lsn.camptocamp.com>
Message-ID: <4DC7B887.7060507@fansubcode.org>

Am 09.05.2011 09:06, schrieb Cedric Jeanneret:
> On Mon, 9 May 2011 14:18:09 +0800
> Jeff Pang<jeffrpang at gmail.com>  wrote:
>
>> Hello,
>>
>> Since HTTPS doesn't support vhost, shall we have the need to put the
>> server_name command in the server section? for example,
>>
>>           server {
>>                  listen       443;
>>                  server_name  xxx.example.com;
>>                  ....
>>           }
>>
>> Thanks.
>>
>>
> o_O since when does HTTPS not support vhosts ?
>
> the only problem is when having multiple SSL vhosts on the same IP - the certificate has to cover all those vhosts.
>
> To answer your question, you shall provide:
> server_name $_;
>
> it's the default vhost
>
> Cheers,
>
> C.

Even this is not needed since nginx does support SNI.

http://en.wikipedia.org/wiki/Server_Name_Indication

Hope this helps a bit,

Pascal


From nginx-forum at nginx.us  Mon May  9 14:52:28 2011
From: nginx-forum at nginx.us (davidkazuhiro)
Date: Mon, 09 May 2011 06:52:28 -0400
Subject: Logs in wrong order
Message-ID: <f94a8df49d13ac9fcd9c4555953b8a64.NginxMailingListEnglish@forum.nginx.org>

When I hit the nginx server with a very rapid succession of requests
from multiple clients, the logs will be out of order. For example, the
log times would appear like this:

... 18:10:11 ...
... 18:10:11 ...
... 18:10:11 ...
... 18:10:12 ...
... 18:10:11 ...
... 18:10:12 ...
... 18:10:11 ...
... 18:10:12 ...
... 18:10:12 ...
... 18:10:12 ...

Why would this be?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196690,196690#msg-196690



From mdounin at mdounin.ru  Mon May  9 14:52:35 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 9 May 2011 14:52:35 +0400
Subject: nginx silently drops response body of POST depending on its length
In-Reply-To: <221afc752b74444b2aa51a8b264f930c.NginxMailingListEnglish@forum.nginx.org>
References: <221afc752b74444b2aa51a8b264f930c.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110509105235.GE42265@mdounin.ru>

Hello!

On Sun, May 08, 2011 at 03:15:28PM -0400, zildjohn01 wrote:

> I've come across an incredibly weird problem with POST responses using
> nginx 1.0.1 and uWSGI 0.9.7.2 on CentOS 5.6.
> 
> uWSGI successfully sends the response body to nginx, but nginx seems to
> ignore it if both (1) the request method is POST, and (2) the response
> length is less than 4052 bytes. The issue doesn't occur when serving
> directly using `uwsgi --http`, so I'm pretty sure the problem is
> something to do with nginx.
> 
> To explain things in code, this Python results in the desired response:
> 
> ....def app_uwsgi(environ, start_response):
> ........start_response('200 OK', [('Content-type', 'text/html')])
> ........return [b'x' * 4052]
> 
> while this results in zero bytes being sent to the client for POST
> requests:
> 
> ....def app_uwsgi(environ, start_response):
> ........start_response('200 OK', [('Content-type', 'text/html')])
> ........return [b'x' * 4051]
> 
> I stripped down to a minimal nginx.conf, and the issue still occurs:
> 
> ....events {
> ........worker_connections  1024;
> ....}
> 
> ....http {
> ........upstream uwsgi_connect {
> ............server  127.0.0.1:39275;
> ........}
> 
> ........server {
> ............listen       20101;
> ............server_name  domain.net;
> 
> ............location / {
> ................uwsgi_pass  uwsgi_connect;
> ................include     uwsgi_params;
> ............}
> ........}
> ....}
> 
> Does anyone have any clue what could be causing this behavior?

As it was already said - you have to read body of POST before 
sending reply.  Else you'll get something like

... readv() failed (54: Connection reset by peer) while reading 
upstream ...

in nginx error log and possibly incorrect reply (depending on various 
factors, including sizes and event method used).

Maxim Dounin


From jeffrpang at gmail.com  Mon May  9 14:58:12 2011
From: jeffrpang at gmail.com (Jeff Pang)
Date: Mon, 9 May 2011 18:58:12 +0800
Subject: log rotate
Message-ID: <BANLkTinzwmdbbo71GNn6-f3d2Sc=mV6h7A@mail.gmail.com>

Hi,

Does nginx support the feature of logrotate by log4j format?
log4j is rotate the log files by each date and rename the old logs
with log_name.YYYY-MM-DD.

Thanks.

-- 
Jeff Pang
www.DNSbed.com


From mdounin at mdounin.ru  Mon May  9 15:08:15 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 9 May 2011 15:08:15 +0400
Subject: Logs in wrong order
In-Reply-To: <f94a8df49d13ac9fcd9c4555953b8a64.NginxMailingListEnglish@forum.nginx.org>
References: <f94a8df49d13ac9fcd9c4555953b8a64.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110509110815.GG42265@mdounin.ru>

Hello!

On Mon, May 09, 2011 at 06:52:28AM -0400, davidkazuhiro wrote:

> When I hit the nginx server with a very rapid succession of requests
> from multiple clients, the logs will be out of order. For example, the
> log times would appear like this:
> 
> ... 18:10:11 ...
> ... 18:10:11 ...
> ... 18:10:11 ...
> ... 18:10:12 ...
> ... 18:10:11 ...
> ... 18:10:12 ...
> ... 18:10:11 ...
> ... 18:10:12 ...
> ... 18:10:12 ...
> ... 18:10:12 ...
> 
> Why would this be?

Time in different nginx worker processes may be slightly different 
as nginx only updates time once per event loop iteration (or even 
less frequently if timer_resolution set).  This in turn results in 
out-of-order times in logs you see when requests from different 
workers are logged.

Similar effect may be also seen when using log buffering (though 
in this case you'll see blocks with successive times up to log 
buffer size, not individual lines).

Maxim Dounin


From mdounin at mdounin.ru  Mon May  9 15:11:24 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 9 May 2011 15:11:24 +0400
Subject: log rotate
In-Reply-To: <BANLkTinzwmdbbo71GNn6-f3d2Sc=mV6h7A@mail.gmail.com>
References: <BANLkTinzwmdbbo71GNn6-f3d2Sc=mV6h7A@mail.gmail.com>
Message-ID: <20110509111124.GH42265@mdounin.ru>

Hello!

On Mon, May 09, 2011 at 06:58:12PM +0800, Jeff Pang wrote:

> Hi,
> 
> Does nginx support the feature of logrotate by log4j format?
> log4j is rotate the log files by each date and rename the old logs
> with log_name.YYYY-MM-DD.

nginx doesn't do any log rotation by itself.  You have to rename 
logs yourself (or with your favorite logrotate tool) and send USR1 
signal to nginx to instruct it to reopen log files.

Maxim Dounin


From mdounin at mdounin.ru  Mon May  9 15:21:49 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 9 May 2011 15:21:49 +0400
Subject: aio info
In-Reply-To: <EFC5A19DBF3A2E489904733B467D990B0F4772@office-server2.m247.local>
References: <EFC5A19DBF3A2E489904733B467D990B0EC5BE@office-server2.m247.local>
	<20110505071212.GA36559@sysoev.ru>
	<EFC5A19DBF3A2E489904733B467D990B0EDC37@office-server2.m247.local>
	<20110506105017.GD3100@sysoev.ru>
	<EFC5A19DBF3A2E489904733B467D990B0EDE0D@office-server2.m247.local>
	<20110506130030.GE3100@sysoev.ru>
	<EFC5A19DBF3A2E489904733B467D990B0F4772@office-server2.m247.local>
Message-ID: <20110509112149.GI42265@mdounin.ru>

Hello!

On Mon, May 09, 2011 at 09:33:18AM +0000, Richard Kearsley wrote:

> Running nginx on freebsd 8.2 with 16 workers and aio on, when the server encounters heavy writes, all processes are in D state
> Aio queue is way below system/process limit
> A request for small file (said file will be in fs buffer/cache) every second varies wildly from 1ms to over 1s
> 
> >From man ps:
> D       Marks a process in disk (or other short term, uninter-ruptible) wait.
> 
> What's causing the lock?

You may want to look at WCHAN instead, it should contain more 
information.

Please note that AIO interface only allows to do async reads, but 
open()/stat()/... are still blocking.  This may be a problem if 
you've run out of vnodes (sysctl kern.maxvnodes) or there are 
other reasons NAMEI lookups can't be effectively cached by OS.

Maxim Dounin


From nginx-forum at nginx.us  Mon May  9 16:13:39 2011
From: nginx-forum at nginx.us (davidkazuhiro)
Date: Mon, 09 May 2011 08:13:39 -0400
Subject: Logs in wrong order
In-Reply-To: <20110509110815.GG42265@mdounin.ru>
References: <20110509110815.GG42265@mdounin.ru>
Message-ID: <0aaee937b2ca97d2ea64d4786bdb7ac8.NginxMailingListEnglish@forum.nginx.org>

Oh I see, tricky. How is an event loop defined? Is there a value
definable in nginx.conf which can shorten the event loop? Or can I just
increase timer resolution to a really high number to make the log times
more accurate?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196690,196707#msg-196707



From lists at ruby-forum.com  Mon May  9 18:17:25 2011
From: lists at ruby-forum.com (Maik Unger)
Date: Mon, 09 May 2011 16:17:25 +0200
Subject: Expierence nGinx with Typo3 Caching
Message-ID: <9eb8c174d8244c592e22be2170d99653@ruby-forum.com>

Hello Everybody,

Does anybody have expierence with nginx and typo3 caching?

I have a very strange problem with the caching on of our customers
sites. They made changes in typo3. On the first domain first.domain.de
the changes will take their effect. When the customer will make another
changes, then the nginx will show the old cache.

What is here the problem? Maybe the proxy_cache_key?

Best Regards,


Maik

-- 
Posted via http://www.ruby-forum.com/.


From lists at ruby-forum.com  Mon May  9 19:53:07 2011
From: lists at ruby-forum.com (John Travolota)
Date: Mon, 09 May 2011 17:53:07 +0200
Subject: How to redirect non anonymous proxy traffic?
Message-ID: <8078670c1a98c6521834d738a287afce@ruby-forum.com>

I know it's something with HTTP_X_FORWARDED directive but not sure how
to make it work.
I need detection of it and then redirect to set URL.

-- 
Posted via http://www.ruby-forum.com/.


From nginx-forum at nginx.us  Mon May  9 20:19:52 2011
From: nginx-forum at nginx.us (zildjohn01)
Date: Mon, 09 May 2011 12:19:52 -0400
Subject: nginx silently drops response body of POST depending on its length
In-Reply-To: <221afc752b74444b2aa51a8b264f930c.NginxMailingListEnglish@forum.nginx.org>
References: <221afc752b74444b2aa51a8b264f930c.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <6982d9bbdfc2231e83c84ca123ab028e.NginxMailingListEnglish@forum.nginx.org>

Thanks, that helped point me to the issue. It turns out the problem was
that I didn't pass --pep3333-input to uWSGI, so it was converting all
CRLFs to LFs in the request body, so CONTENT_LENGTH didn't match the
data length. Passing that flag to uWSGI made it pass the raw stream to
my app and solved the problem.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196581,196761#msg-196761



From ionathan at gmail.com  Mon May  9 20:42:41 2011
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Mon, 9 May 2011 13:42:41 -0300
Subject: proxy_next_upstream and POSTs
Message-ID: <BANLkTimWD4g4=C1mj61fOWH3Wew5fS2ZxA@mail.gmail.com>

Hi!
Nginx is great and the way it handles upstream and failed server as well! I
would just like to avoid retrying request when they are POSTs, as POSTs are
no idempotent and my backend can't handle duplicated POSTs right now.
So I started to mess around with the code. What I have in mind is to create
a module that will change the behavior of proxy_next_upstream by just
checking that http method != POST.
I believe it supposed to be really simple.
So a couple of questions for those who are into nginx source code:
1. Is there any good article, blog post, etc, that explain how to create and
test a module?
2. Any good advice on how I should do it? I was thinking on adding the
validation somewhere in ngx_http_upstream_test_next (ngx_http_upstream.c),
something like:
if (r->method == NGX_HTTP_POST) {
  return NGX_DECLINED;
}

If this is OK, how could I do this from a module? So when someone compiles
nginx with this module, he'll have this behavior instead of the standard
one?

Sorry if my questions are dumb.

Jonathan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110509/5ab0348b/attachment-0001.html>

From nginx-forum at nginx.us  Mon May  9 21:18:34 2011
From: nginx-forum at nginx.us (yellowbox9)
Date: Mon, 09 May 2011 13:18:34 -0400
Subject: empty reply from server
In-Reply-To: <4cd88446bdeea09b413fae442fa115e7.NginxMailingListEnglish@forum.nginx.org>
References: <4cd88446bdeea09b413fae442fa115e7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <8977fb6f50245565eaa9b33541fa9413.NginxMailingListEnglish@forum.nginx.org>

Nothing is in the vhosts dir.... and I'm not doing anything like :

error_page 404 =200 /error;

location /error {
return 444;
}

Not

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196075,196789#msg-196789



From ionathan at gmail.com  Tue May 10 02:37:36 2011
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Mon, 9 May 2011 19:37:36 -0300
Subject: proxy_next_upstream and POSTs
In-Reply-To: <BANLkTimWD4g4=C1mj61fOWH3Wew5fS2ZxA@mail.gmail.com>
References: <BANLkTimWD4g4=C1mj61fOWH3Wew5fS2ZxA@mail.gmail.com>
Message-ID: <BANLkTi=Ebtv30iWgsf4bYmkSfYbnUtZv0g@mail.gmail.com>

I read some modules out there and could manage to write a custom module that
will do the job. I have some doubts though.
How can I call the original upstream module from my own upstream module? I
don't want to replace it.

I think the magic might go here:
ngx_http_upstream_init_<module name>_peer method. Is that right? Is there
any example that I can use?

Thanks!

Jonathan

On Mon, May 9, 2011 at 1:42 PM, Jonathan Leibiusky <ionathan at gmail.com>wrote:

> Hi!
> Nginx is great and the way it handles upstream and failed server as well! I
> would just like to avoid retrying request when they are POSTs, as POSTs are
> no idempotent and my backend can't handle duplicated POSTs right now.
> So I started to mess around with the code. What I have in mind is to create
> a module that will change the behavior of proxy_next_upstream by just
> checking that http method != POST.
> I believe it supposed to be really simple.
> So a couple of questions for those who are into nginx source code:
> 1. Is there any good article, blog post, etc, that explain how to create
> and test a module?
> 2. Any good advice on how I should do it? I was thinking on adding the
> validation somewhere in ngx_http_upstream_test_next (ngx_http_upstream.c),
> something like:
> if (r->method == NGX_HTTP_POST) {
>   return NGX_DECLINED;
> }
>
> If this is OK, how could I do this from a module? So when someone compiles
> nginx with this module, he'll have this behavior instead of the standard
> one?
>
> Sorry if my questions are dumb.
>
> Jonathan
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110509/7f9d492f/attachment.html>

From member at linkedin.com  Tue May 10 10:46:50 2011
From: member at linkedin.com (=?UTF-8?Q?=E5=AD=A6=E9=B9=8F_=E6=9D=8E_via_LinkedIn?=)
Date: Tue, 10 May 2011 06:46:50 +0000 (UTC)
Subject: Invitation to connect on LinkedIn
Message-ID: <169429782.5059641.1305010010543.JavaMail.app@ela4-bed79.prod>

LinkedIn
------------?? ? requested to add you as a connection on LinkedIn:
------------------------------------------

Evert,

I'd like to add you to my professional network on LinkedIn.

- ??

Accept invitation from ?? ?
http://www.linkedin.com/e/-k9ugzs-gnih9kmm-66/vknzOax-I0d5XpARv228ODrxyY/blk/I2800928867_2/1BpC5vrmRLoRZcjkkZt5YCpnlOt3RApnhMpmdzgmhxrSNBszYOnPsSe3wOej0Me399bR1zkDwTkzBdbPwOd30MdPoQe38LrCBxbOYWrSlI/EML_comm_afe/

View invitation from ?? ?
http://www.linkedin.com/e/-k9ugzs-gnih9kmm-66/vknzOax-I0d5XpARv228ODrxyY/blk/I2800928867_2/39vdPoUe38Vc30UcAALqnpPbOYWrSlI/svi/

------------------------------------------ 
DID YOU KNOW your LinkedIn profile helps you control your public image when people search for you? Setting your profile as public means your LinkedIn profile will come up when people enter your name in leading search engines. Take control of your image! 
http://www.linkedin.com/e/-k9ugzs-gnih9kmm-66/ewp/inv-22/

 
-- 
(c) 2011, LinkedIn Corporation
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110510/d50e770c/attachment.html>

From antoine.bonavita at gmail.com  Tue May 10 11:22:27 2011
From: antoine.bonavita at gmail.com (Antoine Bonavita)
Date: Tue, 10 May 2011 09:22:27 +0200
Subject: proxy_next_upstream and POSTs
In-Reply-To: <BANLkTi=Ebtv30iWgsf4bYmkSfYbnUtZv0g@mail.gmail.com>
References: <BANLkTimWD4g4=C1mj61fOWH3Wew5fS2ZxA@mail.gmail.com>
	<BANLkTi=Ebtv30iWgsf4bYmkSfYbnUtZv0g@mail.gmail.com>
Message-ID: <BANLkTind6FuZ9vKaH_K9AyEgRgWczY2X4g@mail.gmail.com>

Hello,

You might want to have a look at Evan's Miller section on proxying
handlers: http://www.evanmiller.org/nginx-modules-guide.html#proxying
This page is pretty much considered *the* reference for module development.

Enjoy,

Antoine.
--
Low fare on trips to nginx land: visit http://www.nginx-discovery.com

On Tue, May 10, 2011 at 12:37 AM, Jonathan Leibiusky <ionathan at gmail.com> wrote:
> I read some modules out there and could manage to write a custom module that
> will do the job. I have some doubts though.
> How can I call the original upstream module from my own upstream module? I
> don't want to replace it.
>
> I think the magic might go here:
> ngx_http_upstream_init_<module name>_peer method. Is that right? Is there
> any example that I can use?
>
> Thanks!
>
> Jonathan
>
> On Mon, May 9, 2011 at 1:42 PM, Jonathan Leibiusky <ionathan at gmail.com>
> wrote:
>>
>> Hi!
>> Nginx is great and the way it handles upstream and failed server as well!
>> I would just like to avoid retrying request when they are POSTs, as POSTs
>> are no idempotent and my backend can't handle duplicated POSTs right now.
>> So I started to mess around with the code. What I have in mind is to
>> create a module that will change the behavior of proxy_next_upstream by just
>> checking that http method != POST.
>> I believe it supposed to be really simple.
>> So a couple of questions for those who are into nginx source code:
>> 1. Is there any good article, blog post, etc, that explain how to create
>> and test a module?
>> 2. Any good advice on how I should do it? I was thinking on adding the
>> validation somewhere in ngx_http_upstream_test_next (ngx_http_upstream.c),
>> something like:
>> if (r->method == NGX_HTTP_POST) {
>> ? return NGX_DECLINED;
>> }
>>
>> If this is OK, how could I do this from a module? So when someone compiles
>> nginx with this module, he'll have this behavior instead of the standard
>> one?
>>
>> Sorry if my questions are dumb.
>>
>> Jonathan
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>


From grails at jmsd.co.uk  Tue May 10 13:31:06 2011
From: grails at jmsd.co.uk (John Moore)
Date: Tue, 10 May 2011 10:31:06 +0100
Subject: 2048 worker_connections are not enough while connecting to upstream
Message-ID: <4DC905DA.7000200@jmsd.co.uk>

I'm using nginx as a remote proxy server in front of some Tomcat 
instances, with pretty high traffic. Suddenly yesterday afternoon we 
started getting the error in the message subject, one I'd never seen 
before, and it's occurred tens of thousands of times since. Now, I've 
upped the worker_connections count to a much higher figure and the 
problem has gone away for now. But I'm concerned that it might be 
indicative of something else - some kind of DOS attack, maybe? What does 
this signify and why might we suddenly start getting it? Thanks.

John


From mdounin at mdounin.ru  Tue May 10 14:19:22 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 10 May 2011 14:19:22 +0400
Subject: 2048 worker_connections are not enough while connecting to
	upstream
In-Reply-To: <4DC905DA.7000200@jmsd.co.uk>
References: <4DC905DA.7000200@jmsd.co.uk>
Message-ID: <20110510101922.GJ42265@mdounin.ru>

Hello!

On Tue, May 10, 2011 at 10:31:06AM +0100, John Moore wrote:

> I'm using nginx as a remote proxy server in front of some Tomcat
> instances, with pretty high traffic. Suddenly yesterday afternoon we
> started getting the error in the message subject, one I'd never seen
> before, and it's occurred tens of thousands of times since. Now,
> I've upped the worker_connections count to a much higher figure and
> the problem has gone away for now. But I'm concerned that it might
> be indicative of something else - some kind of DOS attack, maybe?
> What does this signify and why might we suddenly start getting it?
> Thanks.

Without proper history graphs (use stub_status, Luke!) it's hard 
to say if you are under attack or just your normal load reached 
the bar (or your backend response time degraded and this resulted 
in more connections used, or your network uplink has problems and 
this caused more connections...).

On the other hand, 2048 is pretty low, and easily reached even 
without keepalive.  Usual production values for "high traffic" 
sites is over 9000.

Please also note that using high worker_connections may also 
require tuning of your OS to allow appropriate number of file 
descriptors to be used.

Maxim Dounin


From grails at jmsd.co.uk  Tue May 10 14:32:29 2011
From: grails at jmsd.co.uk (John Moore)
Date: Tue, 10 May 2011 11:32:29 +0100
Subject: 2048 worker_connections are not enough while connecting
	to	upstream
In-Reply-To: <20110510101922.GJ42265@mdounin.ru>
References: <4DC905DA.7000200@jmsd.co.uk> <20110510101922.GJ42265@mdounin.ru>
Message-ID: <4DC9143D.3070903@jmsd.co.uk>

On 10/05/11 11:19, Maxim Dounin wrote:
> Hello!
>
> On Tue, May 10, 2011 at 10:31:06AM +0100, John Moore wrote:
>
>    
>> I'm using nginx as a remote proxy server in front of some Tomcat
>> instances, with pretty high traffic. Suddenly yesterday afternoon we
>> started getting the error in the message subject, one I'd never seen
>> before, and it's occurred tens of thousands of times since. Now,
>> I've upped the worker_connections count to a much higher figure and
>> the problem has gone away for now. But I'm concerned that it might
>> be indicative of something else - some kind of DOS attack, maybe?
>> What does this signify and why might we suddenly start getting it?
>> Thanks.
>>      
> Without proper history graphs (use stub_status, Luke!) it's hard
> to say if you are under attack or just your normal load reached
> the bar (or your backend response time degraded and this resulted
> in more connections used, or your network uplink has problems and
> this caused more connections...).
>    
I'll look into these.
> On the other hand, 2048 is pretty low, and easily reached even
> without keepalive.  Usual production values for "high traffic"
> sites is over 9000.
>    
OK, I may bump it up a bit higher than the 8192 I set it as today.
> Please also note that using high worker_connections may also
> require tuning of your OS to allow appropriate number of file
> descriptors to be used.
>    
I made this change a while back, so we should be OK with that.

Thanks for your extremely lucid help!

John



From igor at sysoev.ru  Tue May 10 16:29:51 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 10 May 2011 16:29:51 +0400
Subject: nginx-1.0.2
Message-ID: <20110510122951.GE90700@sysoev.ru>

Changes with nginx 1.0.2                                         10 May 2011

    *) Feature: now shared zones and caches use POSIX semaphores.

    *) Bugfix: in the "rotate" parameter of the "image_filter" directive. 
       Thanks to Adam Bocim.

    *) Bugfix: nginx could not be built on Solaris; the bug had appeared in 
       1.0.1.


-- 
Igor Sysoev


From ionathan at gmail.com  Tue May 10 18:09:45 2011
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Tue, 10 May 2011 11:09:45 -0300
Subject: proxy_next_upstream and POSTs
In-Reply-To: <BANLkTind6FuZ9vKaH_K9AyEgRgWczY2X4g@mail.gmail.com>
References: <BANLkTimWD4g4=C1mj61fOWH3Wew5fS2ZxA@mail.gmail.com>
	<BANLkTi=Ebtv30iWgsf4bYmkSfYbnUtZv0g@mail.gmail.com>
	<BANLkTind6FuZ9vKaH_K9AyEgRgWczY2X4g@mail.gmail.com>
Message-ID: <BANLkTinQtu2pCrofthkSqA0hZwDKScPmGw@mail.gmail.com>

Thanks! That guide was really useful.
I almost have it done. What is missing there is how to test and maybe
something even more generic.
Once I finish my module I will try to write an even more basic guide
to help people :)

Jonathan

On 5/10/11, Antoine Bonavita <antoine.bonavita at gmail.com> wrote:
> Hello,
>
> You might want to have a look at Evan's Miller section on proxying
> handlers: http://www.evanmiller.org/nginx-modules-guide.html#proxying
> This page is pretty much considered *the* reference for module development.
>
> Enjoy,
>
> Antoine.
> --
> Low fare on trips to nginx land: visit http://www.nginx-discovery.com
>
> On Tue, May 10, 2011 at 12:37 AM, Jonathan Leibiusky <ionathan at gmail.com>
> wrote:
>> I read some modules out there and could manage to write a custom module
>> that
>> will do the job. I have some doubts though.
>> How can I call the original upstream module from my own upstream module? I
>> don't want to replace it.
>>
>> I think the magic might go here:
>> ngx_http_upstream_init_<module name>_peer method. Is that right? Is there
>> any example that I can use?
>>
>> Thanks!
>>
>> Jonathan
>>
>> On Mon, May 9, 2011 at 1:42 PM, Jonathan Leibiusky <ionathan at gmail.com>
>> wrote:
>>>
>>> Hi!
>>> Nginx is great and the way it handles upstream and failed server as well!
>>> I would just like to avoid retrying request when they are POSTs, as POSTs
>>> are no idempotent and my backend can't handle duplicated POSTs right now.
>>> So I started to mess around with the code. What I have in mind is to
>>> create a module that will change the behavior of proxy_next_upstream by
>>> just
>>> checking that http method != POST.
>>> I believe it supposed to be really simple.
>>> So a couple of questions for those who are into nginx source code:
>>> 1. Is there any good article, blog post, etc, that explain how to create
>>> and test a module?
>>> 2. Any good advice on how I should do it? I was thinking on adding the
>>> validation somewhere in ngx_http_upstream_test_next
>>> (ngx_http_upstream.c),
>>> something like:
>>> if (r->method == NGX_HTTP_POST) {
>>> ? return NGX_DECLINED;
>>> }
>>>
>>> If this is OK, how could I do this from a module? So when someone
>>> compiles
>>> nginx with this module, he'll have this behavior instead of the standard
>>> one?
>>>
>>> Sorry if my questions are dumb.
>>>
>>> Jonathan
>>
>>
>> _______________________________________________
>> nginx mailing list
>> nginx at nginx.org
>> http://nginx.org/mailman/listinfo/nginx
>>
>>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>


From piotr.sikora at frickle.com  Tue May 10 18:32:08 2011
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Tue, 10 May 2011 16:32:08 +0200
Subject: proxy_next_upstream and POSTs
In-Reply-To: <BANLkTi=Ebtv30iWgsf4bYmkSfYbnUtZv0g@mail.gmail.com>
References: <BANLkTimWD4g4=C1mj61fOWH3Wew5fS2ZxA@mail.gmail.com>
	<BANLkTi=Ebtv30iWgsf4bYmkSfYbnUtZv0g@mail.gmail.com>
Message-ID: <FDA7BA958BED4B28855060876BE0BB51@Desktop>

Hi,

> 2. Any good advice on how I should do it? I was thinking on adding the 
> validation somewhere in ngx_http_upstream_test_next (ngx_http_upstream.c), 
> something like:
> if (r->method == NGX_HTTP_POST) {
>   return NGX_DECLINED;
> }

Keep in mind that you should decline sending request to another backend, 
only if you already sent it upstream. There is a bunch of errors (bad 
gateway, connection timeout, etc) which could lead to the code path in 
question, for which you should still send request to another backend.

Best regards,
Piotr Sikora < piotr.sikora at frickle.com >



From ionathan at gmail.com  Tue May 10 18:48:28 2011
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Tue, 10 May 2011 11:48:28 -0300
Subject: proxy_next_upstream and POSTs
In-Reply-To: <FDA7BA958BED4B28855060876BE0BB51@Desktop>
References: <BANLkTimWD4g4=C1mj61fOWH3Wew5fS2ZxA@mail.gmail.com>
	<BANLkTi=Ebtv30iWgsf4bYmkSfYbnUtZv0g@mail.gmail.com>
	<FDA7BA958BED4B28855060876BE0BB51@Desktop>
Message-ID: <BANLkTinLb82m8+FzHdsCUWM=2trmyBgPmA@mail.gmail.com>

Right, that is why after digging for a while I realized I wan't doing it the
right way.
Now I wrote a module that will hook into the upstream logic and will set all
the callbacks to the round-robin callbacks. Only the the peer.free callback
will call the original round-robin callback and after that will do something
like:

if (r->method == NGX_HTTP_POST) {
  pc->retries = 0;
}

by setting retires = 0 I think nginx won't retry and will fail and I will be
using round-robin logic for all the rest.

My only problem is that in the peer.free callback I don't have access to the
request struct, so I don't know is the request's method. So now I am not
sure if the approach is the right one.

Any thoughts on this?

On Tue, May 10, 2011 at 11:32 AM, Piotr Sikora <piotr.sikora at frickle.com>wrote:

> Hi,
>
>
>  2. Any good advice on how I should do it? I was thinking on adding the
>> validation somewhere in ngx_http_upstream_test_next (ngx_http_upstream.c),
>> something like:
>> if (r->method == NGX_HTTP_POST) {
>>  return NGX_DECLINED;
>> }
>>
>
> Keep in mind that you should decline sending request to another backend,
> only if you already sent it upstream. There is a bunch of errors (bad
> gateway, connection timeout, etc) which could lead to the code path in
> question, for which you should still send request to another backend.
>
> Best regards,
> Piotr Sikora < piotr.sikora at frickle.com >
>
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110510/213bdeb1/attachment.html>

From ionathan at gmail.com  Tue May 10 23:30:30 2011
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Tue, 10 May 2011 16:30:30 -0300
Subject: proxy_next_upstream and POSTs
In-Reply-To: <BANLkTinLb82m8+FzHdsCUWM=2trmyBgPmA@mail.gmail.com>
References: <BANLkTimWD4g4=C1mj61fOWH3Wew5fS2ZxA@mail.gmail.com>
	<BANLkTi=Ebtv30iWgsf4bYmkSfYbnUtZv0g@mail.gmail.com>
	<FDA7BA958BED4B28855060876BE0BB51@Desktop>
	<BANLkTinLb82m8+FzHdsCUWM=2trmyBgPmA@mail.gmail.com>
Message-ID: <BANLkTik-SVfT=iDfH8hNh_aGct_1d58Sow@mail.gmail.com>

Ok. Finally I could work it out, so thanks anyway!

Here is my small contribution to nginx, it is not much, but it is something.

https://github.com/xetorthio/nginx-upstream-idempotent

Thanks everyone for your help, and if you have time and find something
weired in the code, let me know!

Jonathan

On Tue, May 10, 2011 at 11:48 AM, Jonathan Leibiusky <ionathan at gmail.com>wrote:

> Right, that is why after digging for a while I realized I wan't doing it
> the right way.
> Now I wrote a module that will hook into the upstream logic and will set
> all the callbacks to the round-robin callbacks. Only the the peer.free
> callback will call the original round-robin callback and after that will do
> something like:
>
>
> if (r->method == NGX_HTTP_POST) {
>   pc->retries = 0;
> }
>
> by setting retires = 0 I think nginx won't retry and will fail and I will
> be using round-robin logic for all the rest.
>
> My only problem is that in the peer.free callback I don't have access to
> the request struct, so I don't know is the request's method. So now I am not
> sure if the approach is the right one.
>
> Any thoughts on this?
>
>
> On Tue, May 10, 2011 at 11:32 AM, Piotr Sikora <piotr.sikora at frickle.com>wrote:
>
>> Hi,
>>
>>
>>  2. Any good advice on how I should do it? I was thinking on adding the
>>> validation somewhere in ngx_http_upstream_test_next (ngx_http_upstream.c),
>>> something like:
>>> if (r->method == NGX_HTTP_POST) {
>>>  return NGX_DECLINED;
>>> }
>>>
>>
>> Keep in mind that you should decline sending request to another backend,
>> only if you already sent it upstream. There is a bunch of errors (bad
>> gateway, connection timeout, etc) which could lead to the code path in
>> question, for which you should still send request to another backend.
>>
>> Best regards,
>> Piotr Sikora < piotr.sikora at frickle.com >
>>
>>
>>
>> _______________________________________________
>> nginx mailing list
>> nginx at nginx.org
>> http://nginx.org/mailman/listinfo/nginx
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110510/128bab05/attachment-0001.html>

From nginx-forum at nginx.us  Wed May 11 01:59:41 2011
From: nginx-forum at nginx.us (jeanjan)
Date: Tue, 10 May 2011 17:59:41 -0400
Subject: cgi files in plain text
Message-ID: <875369c24424bd6f71a4658c85bce47b.NginxMailingListEnglish@forum.nginx.org>

Hi,

I installed nginx and php5-fpm from repositories on ubuntu 10.04.
everything works well. The phpinfo test file works.

I try to access collection (for collectd) pages. When I access the url
of the collection3/bin/index.cgi, I can see the file in plain text.

This is the config file used :
server {
    listen   80;
    server_name  toto.com;
    access_log  /var/log/nginx/localhost.access.log;

## Default location
    location / {
        root   /var/www;
        index  index.php;
	try_files $uri $uri/ /index.php?q=$uri&$args;
	auth_basic            "Restricted";
	auth_basic_user_file  /var/htpasswd_au1;

    }

# Protection
#location = /var/www {
#  auth_basic            "Restricted";
#  auth_basic_user_file  /var/htpasswd_au1;
#}

##    try_files $uri $uri/ /index.php?q=$uri;
## Images and static content is treated different
    location ~* ^.+.(jpg|jpeg|gif|css|png|js|ico|xml)$ {
      access_log        off;
      expires           30d;
      root /var/www;
    }



## Parse all .php file in the /var/www directory
    location ~ .php$ {
        fastcgi_split_path_info ^(.+\.php)(.*)$;
        fastcgi_pass   127.0.0.1:9000;
        fastcgi_index  index.php;
        fastcgi_param  SCRIPT_FILENAME  /var/www$fastcgi_script_name;
        include fastcgi_params;
        fastcgi_param  QUERY_STRING     $query_string;
        fastcgi_param  REQUEST_METHOD   $request_method;
        fastcgi_param  CONTENT_TYPE     $content_type;
        fastcgi_param  CONTENT_LENGTH   $content_length;
        fastcgi_intercept_errors        on;
        fastcgi_ignore_client_abort     off;
        fastcgi_connect_timeout 60;
        fastcgi_send_timeout 180;
        fastcgi_read_timeout 180;
        fastcgi_buffer_size 128k;
        fastcgi_buffers 4 256k;
        fastcgi_busy_buffers_size 256k;
        fastcgi_temp_file_write_size 256k;
    }

## Disable viewing .htaccess & .htpassword
    location ~ /\.ht {
        deny  all;
    }
}
upstream backend {
        server 127.0.0.1:9000;
}

Thanks

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197100,197100#msg-197100



From nginx-forum at nginx.us  Wed May 11 03:07:36 2011
From: nginx-forum at nginx.us (brianmercer)
Date: Tue, 10 May 2011 19:07:36 -0400
Subject: cgi files in plain text
In-Reply-To: <875369c24424bd6f71a4658c85bce47b.NginxMailingListEnglish@forum.nginx.org>
References: <875369c24424bd6f71a4658c85bce47b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <fdfb573448785f1c1c291a933c8ea267.NginxMailingListEnglish@forum.nginx.org>

jeanjan Wrote:
-------------------------------------------------------
> Hi,
> 
> I installed nginx and php5-fpm from repositories
> on ubuntu 10.04.
> everything works well. The phpinfo test file
> works.
> 
> I try to access collection (for collectd) pages.
> When I access the url of the
> collection3/bin/index.cgi, I can see the file in
> plain text.
> 
> This is the config file used :
> server {
>     listen   80;
>     server_name  toto.com;
>     access_log 
> /var/log/nginx/localhost.access.log;
> 
> ## Default location
>     location / {
>         root   /var/www;
>         index  index.php;
> 	try_files $uri $uri/ /index.php?q=$uri&$args;
> 	auth_basic            "Restricted";
> 	auth_basic_user_file  /var/htpasswd_au1;
> 
>     }
> 
> # Protection
> #location = /var/www {
> #  auth_basic            "Restricted";
> #  auth_basic_user_file  /var/htpasswd_au1;
> #}
> 
> ##    try_files $uri $uri/ /index.php?q=$uri;
> ## Images and static content is treated different
>     location ~*
> ^.+.(jpg|jpeg|gif|css|png|js|ico|xml)$ {
>       access_log        off;
>       expires           30d;
>       root /var/www;
>     }
> 
> 
> 
> ## Parse all .php file in the /var/www directory
>     location ~ .php$ {
>         fastcgi_split_path_info ^(.+\.php)(.*)$;
>         fastcgi_pass   127.0.0.1:9000;
>         fastcgi_index  index.php;
>         fastcgi_param  SCRIPT_FILENAME 
> /var/www$fastcgi_script_name;
>         include fastcgi_params;
>         fastcgi_param  QUERY_STRING    
> $query_string;
>         fastcgi_param  REQUEST_METHOD  
> $request_method;
>         fastcgi_param  CONTENT_TYPE    
> $content_type;
>         fastcgi_param  CONTENT_LENGTH  
> $content_length;
>         fastcgi_intercept_errors        on;
>         fastcgi_ignore_client_abort     off;
>         fastcgi_connect_timeout 60;
>         fastcgi_send_timeout 180;
>         fastcgi_read_timeout 180;
>         fastcgi_buffer_size 128k;
>         fastcgi_buffers 4 256k;
>         fastcgi_busy_buffers_size 256k;
>         fastcgi_temp_file_write_size 256k;
>     }
> 
> ## Disable viewing .htaccess & .htpassword
>     location ~ /\.ht {
>         deny  all;
>     }
> }
> upstream backend {
>         server 127.0.0.1:9000;
> }
> 
> Thanks

You can either use a php-based web interface for collectd like
php-collection or Collectd Graph Panel or if you need the ability to
serve .cgi and .pl files (like for awstats) you'll have to install
another backend daemon.

You can use a light web server or I'd recommend fcgiwrap.  A couple ppas
have Lucid backports of the Ubuntu fcgiwrap package, (e.g.
https://launchpad.net/~wwu-numerik/+archive/lucid).

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197100,197123#msg-197123



From antoine.bonavita at gmail.com  Wed May 11 12:03:10 2011
From: antoine.bonavita at gmail.com (Antoine Bonavita)
Date: Wed, 11 May 2011 10:03:10 +0200
Subject: proxy_next_upstream and POSTs
In-Reply-To: <BANLkTik-SVfT=iDfH8hNh_aGct_1d58Sow@mail.gmail.com>
References: <BANLkTimWD4g4=C1mj61fOWH3Wew5fS2ZxA@mail.gmail.com>
	<BANLkTi=Ebtv30iWgsf4bYmkSfYbnUtZv0g@mail.gmail.com>
	<FDA7BA958BED4B28855060876BE0BB51@Desktop>
	<BANLkTinLb82m8+FzHdsCUWM=2trmyBgPmA@mail.gmail.com>
	<BANLkTik-SVfT=iDfH8hNh_aGct_1d58Sow@mail.gmail.com>
Message-ID: <BANLkTinjSiT3rJJ+JhGsVHMT_xRfnw2t5w@mail.gmail.com>

Glad to hear you made it.
As far as testing goes, you might find interesting (or just fun) to
read my own experiences with the subject:
- http://www.nginx-discovery.com/2011/02/day-29-nginx-development-tool-chain.html
- http://www.nginx-discovery.com/2011/03/day-32-moving-to-testnginx.html
- http://www.nginx-discovery.com/2011/03/day-33-testnginx-pipelinedrequests.html
- http://www.nginx-discovery.com/2011/03/day-40-testnginx-new-features.html

As far as documentation of Test::Nginx goes, you might find the devel
tree (https://github.com/agentzh/test-nginx/tree/devel) which includes
my commits a bit more verbose than the "master" branch.

Enjoy testing.

A.

--
Round-trip fares to nginx land as low as $0: visit
http://www.nginx-discovery.com

On Tue, May 10, 2011 at 9:30 PM, Jonathan Leibiusky <ionathan at gmail.com> wrote:
> Ok. Finally I could work it out, so thanks anyway!
>
> Here is my small contribution to nginx, it is not much, but it is something.
>
> https://github.com/xetorthio/nginx-upstream-idempotent
>
> Thanks everyone for your help, and if you have time and find something
> weired in the code, let me know!
>
> Jonathan
>
> On Tue, May 10, 2011 at 11:48 AM, Jonathan Leibiusky <ionathan at gmail.com>
> wrote:
>>
>> Right, that is why after digging for a while I realized I wan't doing it
>> the right way.
>> Now I wrote a module that will hook into the upstream logic and will set
>> all the callbacks to the round-robin callbacks. Only the the peer.free
>> callback will call the original round-robin callback and after that will do
>> something like:
>>
>> if (r->method == NGX_HTTP_POST) {
>> ? pc->retries = 0;
>> }
>>
>> by setting retires = 0 I think nginx won't retry and will fail and I will
>> be using round-robin logic for all the rest.
>>
>> My only problem is that in the peer.free callback I don't have access to
>> the request struct, so I don't know is the request's method. So now I am not
>> sure if the approach is the right one.
>>
>> Any thoughts on this?
>>
>> On Tue, May 10, 2011 at 11:32 AM, Piotr Sikora <piotr.sikora at frickle.com>
>> wrote:
>>>
>>> Hi,
>>>
>>>> 2. Any good advice on how I should do it? I was thinking on adding the
>>>> validation somewhere in ngx_http_upstream_test_next (ngx_http_upstream.c),
>>>> something like:
>>>> if (r->method == NGX_HTTP_POST) {
>>>> ?return NGX_DECLINED;
>>>> }
>>>
>>> Keep in mind that you should decline sending request to another backend,
>>> only if you already sent it upstream. There is a bunch of errors (bad
>>> gateway, connection timeout, etc) which could lead to the code path in
>>> question, for which you should still send request to another backend.
>>>
>>> Best regards,
>>> Piotr Sikora < piotr.sikora at frickle.com >
>>>
>>>
>>> _______________________________________________
>>> nginx mailing list
>>> nginx at nginx.org
>>> http://nginx.org/mailman/listinfo/nginx
>>
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>


From igor at sysoev.ru  Wed May 11 14:06:16 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 11 May 2011 14:06:16 +0400
Subject: server_name for https
In-Reply-To: <20110509090627.68905d0d@cholatse.wrk.lsn.camptocamp.com>
References: <BANLkTim=GtK0AnJvb_XiTCt6nGmjsNpaTQ@mail.gmail.com>
	<20110509090627.68905d0d@cholatse.wrk.lsn.camptocamp.com>
Message-ID: <20110511100615.GB45105@sysoev.ru>

On Mon, May 09, 2011 at 09:06:27AM +0200, Cedric Jeanneret wrote:
> On Mon, 9 May 2011 14:18:09 +0800
> Jeff Pang <jeffrpang at gmail.com> wrote:
> 
> > Hello,
> > 
> > Since HTTPS doesn't support vhost, shall we have the need to put the
> > server_name command in the server section? for example,
> > 
> >          server {
> >                 listen       443;
> >                 server_name  xxx.example.com;
> >                 ....
> >          }
> > 
> > Thanks.
> > 
> > 
> 
> o_O since when does HTTPS not support vhosts ?
> 
> the only problem is when having multiple SSL vhosts on the same IP - the certificate has to cover all those vhosts.
> 
> To answer your question, you shall provide:
> server_name $_;
> 
> it's the default vhost

This is not default virtual host. Please read this:
http://nginx.org/en/docs/http/server_names.html#miscellaneous_names


-- 
Igor Sysoev


From igor at sysoev.ru  Wed May 11 14:07:20 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 11 May 2011 14:07:20 +0400
Subject: server_name for https
In-Reply-To: <BANLkTim=GtK0AnJvb_XiTCt6nGmjsNpaTQ@mail.gmail.com>
References: <BANLkTim=GtK0AnJvb_XiTCt6nGmjsNpaTQ@mail.gmail.com>
Message-ID: <20110511100720.GC45105@sysoev.ru>

On Mon, May 09, 2011 at 02:18:09PM +0800, Jeff Pang wrote:
> Hello,
> 
> Since HTTPS doesn't support vhost, shall we have the need to put the
> server_name command in the server section? for example,
> 
>          server {
>                 listen       443;
>                 server_name  xxx.example.com;
>                 ....
>          }
> 
> Thanks.

server should always have a name. Or it will be "" by default.


-- 
Igor Sysoev


From cjeanneret at internux.ch  Wed May 11 14:13:23 2011
From: cjeanneret at internux.ch (Cedric Jeanneret)
Date: Wed, 11 May 2011 12:13:23 +0200
Subject: server_name for https
In-Reply-To: <20110511100615.GB45105@sysoev.ru>
References: <BANLkTim=GtK0AnJvb_XiTCt6nGmjsNpaTQ@mail.gmail.com>
	<20110509090627.68905d0d@cholatse.wrk.lsn.camptocamp.com>
	<20110511100615.GB45105@sysoev.ru>
Message-ID: <20110511121323.60180dc7@cholatse.wrk.lsn.camptocamp.com>

On Wed, 11 May 2011 14:06:16 +0400
Igor Sysoev <igor at sysoev.ru> wrote:

> On Mon, May 09, 2011 at 09:06:27AM +0200, Cedric Jeanneret wrote:
> > On Mon, 9 May 2011 14:18:09 +0800
> > Jeff Pang <jeffrpang at gmail.com> wrote:
> > 
> > > Hello,
> > > 
> > > Since HTTPS doesn't support vhost, shall we have the need to put the
> > > server_name command in the server section? for example,
> > > 
> > >          server {
> > >                 listen       443;
> > >                 server_name  xxx.example.com;
> > >                 ....
> > >          }
> > > 
> > > Thanks.
> > > 
> > > 
> > 
> > o_O since when does HTTPS not support vhosts ?
> > 
> > the only problem is when having multiple SSL vhosts on the same IP - the certificate has to cover all those vhosts.
> > 
> > To answer your question, you shall provide:
> > server_name $_;
> > 
> > it's the default vhost
> 
> This is not default virtual host. Please read this:
> http://nginx.org/en/docs/http/server_names.html#miscellaneous_names
> 
> 
Woops, indeed - sorry :/ (never used this syntax in my installation... )
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://nginx.org/pipermail/nginx/attachments/20110511/607871e6/attachment.pgp>

From nginx-forum at nginx.us  Wed May 11 14:27:38 2011
From: nginx-forum at nginx.us (zls)
Date: Wed, 11 May 2011 06:27:38 -0400
Subject: nginx-1.0.2
In-Reply-To: <20110510122951.GE90700@sysoev.ru>
References: <20110510122951.GE90700@sysoev.ru>
Message-ID: <25b6ac0454130aaa48b9dbc133ad36a2.NginxMailingListEnglish@forum.nginx.org>

feature not enabled on redhat by default because of no -pthread on
testing?


checking for POSIX semaphores

/tmp/cc2q2SHU.o: In function `main':
/search/zls/nginx/server_traffic/nginx-1.0.2.zls/objs/autotest.c:8:
undefined reference to `sem_init'
/search/zls/nginx/server_traffic/nginx-1.0.2.zls/objs/autotest.c:9:
undefined reference to `sem_destroy'
collect2: ld return 1

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196951,197225#msg-197225



From igor at sysoev.ru  Wed May 11 14:30:48 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 11 May 2011 14:30:48 +0400
Subject: nginx-1.0.2
In-Reply-To: <25b6ac0454130aaa48b9dbc133ad36a2.NginxMailingListEnglish@forum.nginx.org>
References: <20110510122951.GE90700@sysoev.ru>
	<25b6ac0454130aaa48b9dbc133ad36a2.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110511103048.GG45105@sysoev.ru>

On Wed, May 11, 2011 at 06:27:38AM -0400, zls wrote:
> feature not enabled on redhat by default because of no -pthread on
> testing?

Yes, POSIX semaphores on Linux requires linking with libpthread.

> checking for POSIX semaphores
> 
> /tmp/cc2q2SHU.o: In function `main':
> /search/zls/nginx/server_traffic/nginx-1.0.2.zls/objs/autotest.c:8:
> undefined reference to `sem_init'
> /search/zls/nginx/server_traffic/nginx-1.0.2.zls/objs/autotest.c:9:
> undefined reference to `sem_destroy'
> collect2: ld return 1


-- 
Igor Sysoev


From nginx-forum at nginx.us  Wed May 11 16:23:53 2011
From: nginx-forum at nginx.us (tuffclassified)
Date: Wed, 11 May 2011 08:23:53 -0400
Subject: redirecting url from old to new
Message-ID: <be28476b9c1398f9c40ef7f620998428.NginxMailingListEnglish@forum.nginx.org>

Any body will help me to redirect my old  wordpress post
http://tuffclassified.com/xyz/ to new custom post
http://tuffclassified.com/ads/xyz/  . It was easy in apache as i have to
add just 

# END WordPress
RewriteRule ^/(.*)$ http://tuffclassified.com/ads/$1 [R=301,L]

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197247,197247#msg-197247



From edhoprima at gmail.com  Wed May 11 16:30:43 2011
From: edhoprima at gmail.com (Edho P Arief)
Date: Wed, 11 May 2011 19:30:43 +0700
Subject: redirecting url from old to new
In-Reply-To: <be28476b9c1398f9c40ef7f620998428.NginxMailingListEnglish@forum.nginx.org>
References: <be28476b9c1398f9c40ef7f620998428.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTi=hiHLsZL=zBJDr-AJ8AzRbLYRvyg@mail.gmail.com>

On Wed, May 11, 2011 at 7:23 PM, tuffclassified <nginx-forum at nginx.us> wrote:
> Any body will help me to redirect my old ?wordpress post
> http://tuffclassified.com/xyz/ to new custom post
> http://tuffclassified.com/ads/xyz/ ?. It was easy in apache as i have to
> add just
>
> # END WordPress
> RewriteRule ^/(.*)$ http://tuffclassified.com/ads/$1 [R=301,L]
>

location /xyz/ {
  rewrite ^ /ads/$uri permanent;
  ...
}


From mp3geek at gmail.com  Thu May 12 06:13:03 2011
From: mp3geek at gmail.com (Ryan B)
Date: Thu, 12 May 2011 14:13:03 +1200
Subject: rewrite n00b
Message-ID: <BANLkTin7mX3yD8r1Uwq+mwaW-SMaBCWwYw@mail.gmail.com>

Hi,
Just wanting to enabled a rewrite of the following url,

Redirecting all requests from http://www.domain.com/site/forums and
http://www.domain.com/forums to http://forums.domain.com

Thanks :)


From igor at sysoev.ru  Thu May 12 09:20:53 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Thu, 12 May 2011 09:20:53 +0400
Subject: rewrite n00b
In-Reply-To: <BANLkTin7mX3yD8r1Uwq+mwaW-SMaBCWwYw@mail.gmail.com>
References: <BANLkTin7mX3yD8r1Uwq+mwaW-SMaBCWwYw@mail.gmail.com>
Message-ID: <20110512052052.GC99505@sysoev.ru>

On Thu, May 12, 2011 at 02:13:03PM +1200, Ryan B wrote:
> Hi,
> Just wanting to enabled a rewrite of the following url,
> 
> Redirecting all requests from http://www.domain.com/site/forums and
> http://www.domain.com/forums to http://forums.domain.com

server {
    server_name  www.domain.com;

    location /forums {
        rewrite  ^/forums(.*)$  http://forums.domain.com$1;
    }

    location /site/forums {
        rewrite  ^/site/forums(.*)$  http://forums.domain.com$1;
    }

    ...


-- 
Igor Sysoev


From agentzh at gmail.com  Thu May 12 14:30:52 2011
From: agentzh at gmail.com (agentzh)
Date: Thu, 12 May 2011 18:30:52 +0800
Subject: proxy_next_upstream and POSTs
In-Reply-To: <BANLkTinjSiT3rJJ+JhGsVHMT_xRfnw2t5w@mail.gmail.com>
References: <BANLkTimWD4g4=C1mj61fOWH3Wew5fS2ZxA@mail.gmail.com>
	<BANLkTi=Ebtv30iWgsf4bYmkSfYbnUtZv0g@mail.gmail.com>
	<FDA7BA958BED4B28855060876BE0BB51@Desktop>
	<BANLkTinLb82m8+FzHdsCUWM=2trmyBgPmA@mail.gmail.com>
	<BANLkTik-SVfT=iDfH8hNh_aGct_1d58Sow@mail.gmail.com>
	<BANLkTinjSiT3rJJ+JhGsVHMT_xRfnw2t5w@mail.gmail.com>
Message-ID: <BANLkTingaLBk09hjRU+xvSdEayZ9NfXcdw@mail.gmail.com>

On Wed, May 11, 2011 at 4:03 PM, Antoine Bonavita
<antoine.bonavita at gmail.com> wrote:
> As far as documentation of Test::Nginx goes, you might find the devel
> tree (https://github.com/agentzh/test-nginx/tree/devel) which includes
> my commits a bit more verbose than the "master" branch.
>

I've just merged test-nginx's devel branch back to its master after
massive regression testing on my side. And I'll do a new CPAN release
of Test::Nginx tomorrow or so :)

Happy hacking!
-agentzh


From agentzh at gmail.com  Thu May 12 14:35:28 2011
From: agentzh at gmail.com (agentzh)
Date: Thu, 12 May 2011 18:35:28 +0800
Subject: proxy_next_upstream and POSTs
In-Reply-To: <BANLkTinjSiT3rJJ+JhGsVHMT_xRfnw2t5w@mail.gmail.com>
References: <BANLkTimWD4g4=C1mj61fOWH3Wew5fS2ZxA@mail.gmail.com>
	<BANLkTi=Ebtv30iWgsf4bYmkSfYbnUtZv0g@mail.gmail.com>
	<FDA7BA958BED4B28855060876BE0BB51@Desktop>
	<BANLkTinLb82m8+FzHdsCUWM=2trmyBgPmA@mail.gmail.com>
	<BANLkTik-SVfT=iDfH8hNh_aGct_1d58Sow@mail.gmail.com>
	<BANLkTinjSiT3rJJ+JhGsVHMT_xRfnw2t5w@mail.gmail.com>
Message-ID: <BANLkTimkLKGLNqFpg9yp44xoADkUQqPc-A@mail.gmail.com>

On Wed, May 11, 2011 at 4:03 PM, Antoine Bonavita
<antoine.bonavita at gmail.com> wrote:
> As far as documentation of Test::Nginx goes, you might find the devel
> tree (https://github.com/agentzh/test-nginx/tree/devel) which includes
> my commits a bit more verbose than the "master" branch.
>

I've just merged test-nginx's devel branch back to its master after
massive regression testing on my side. And I'll do a new CPAN release
of Test::Nginx tomorrow or so :)

Happy hacking!
-agentzh


From ash2kk at gmail.com  Thu May 12 15:34:09 2011
From: ash2kk at gmail.com (Mikhail Mazursky)
Date: Thu, 12 May 2011 17:34:09 +0600
Subject: redirecting url from old to new
In-Reply-To: <be28476b9c1398f9c40ef7f620998428.NginxMailingListEnglish@forum.nginx.org>
References: <be28476b9c1398f9c40ef7f620998428.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTiksQH+-9ZAu5YBxmxh7gqX8ALd8cQ@mail.gmail.com>

2011/5/11 tuffclassified <nginx-forum at nginx.us>:
> Any body will help me to redirect my old ?wordpress post
> http://tuffclassified.com/xyz/ to new custom post
> http://tuffclassified.com/ads/xyz/ ?. It was easy in apache as i have to
> add just
>
> # END WordPress
> RewriteRule ^/(.*)$ http://tuffclassified.com/ads/$1 [R=301,L]

location / {
     rewrite ^ /ads$uri permanent;
}


From nginx-forum at nginx.us  Thu May 12 15:58:20 2011
From: nginx-forum at nginx.us (tuffclassified)
Date: Thu, 12 May 2011 07:58:20 -0400
Subject: redirecting url from old to new
In-Reply-To: <BANLkTiksQH+-9ZAu5YBxmxh7gqX8ALd8cQ@mail.gmail.com>
References: <BANLkTiksQH+-9ZAu5YBxmxh7gqX8ALd8cQ@mail.gmail.com>
Message-ID: <ebb704f3810858402ae2df71b2c97a58.NginxMailingListEnglish@forum.nginx.org>

thanks later one seems to work

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197247,197578#msg-197578



From kworthington at gmail.com  Thu May 12 18:06:04 2011
From: kworthington at gmail.com (Kevin Worthington)
Date: Thu, 12 May 2011 10:06:04 -0400
Subject: Nginx-1.0.2 for Windows 32-bit and 64-bit now available
Message-ID: <BANLkTikjmU=FsUORcELUKHbWmgNXnchmLw@mail.gmail.com>

For those interested, Cygwin based versions of Nginx 1.0.2 for Windows
(both 32-bit and 64-bit) builds are available here:
http://www.kevinworthington.com/nginx-for-windows/

Blog post/announcement: http://www.kevinworthington.com/nginx-for-windows-1-0-2/

Cygwin is NOT required on your system to run this version; all needed
files are included.

Please note that I have been providing Cygwin based builds since
before official Windows builds were made available. It's mostly to
support folks already using the builds.

Thank you,
Kevin
--
Kevin Worthington
http://www.kevinworthington.com/


From nginx-forum at nginx.us  Thu May 12 22:25:25 2011
From: nginx-forum at nginx.us (rubix81)
Date: Thu, 12 May 2011 14:25:25 -0400
Subject: Locate Command Linux Problems
Message-ID: <cf7f1ada40ce35dd041bd59d428a9570.NginxMailingListEnglish@forum.nginx.org>

Hey so I am trying to write installation software. I am installing third
party software on a computer, but first I check to see if it has already
been installed. I am trying to use the locate command but it is not
returning anything. To test it, I put a file that I installed and copied
it into a different directory. I know it is there, and I copy and pasted
the title so that I knew that would be correct. "$locate filename.a"
returns nothing, however. I can be sitting in the correct directory
staring straight at the file and it will still not give me its location.
Does anyone know more about how locate works who can help me figure this
out?

Thanks!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197663,197663#msg-197663



From dan.trainor at gmail.com  Thu May 12 22:30:58 2011
From: dan.trainor at gmail.com (Dan Trainor)
Date: Thu, 12 May 2011 11:30:58 -0700
Subject: Locate Command Linux Problems
In-Reply-To: <cf7f1ada40ce35dd041bd59d428a9570.NginxMailingListEnglish@forum.nginx.org>
References: <cf7f1ada40ce35dd041bd59d428a9570.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTi=Y2NFxZRYOpYkO8P45dRhNqR8VuA@mail.gmail.com>

On Thu, May 12, 2011 at 11:25 AM, rubix81 <nginx-forum at nginx.us> wrote:
> Hey so I am trying to write installation software. I am installing third
> party software on a computer, but first I check to see if it has already
> been installed. I am trying to use the locate command but it is not
> returning anything. To test it, I put a file that I installed and copied
> it into a different directory. I know it is there, and I copy and pasted
> the title so that I knew that would be correct. "$locate filename.a"
> returns nothing, however. I can be sitting in the correct directory
> staring straight at the file and it will still not give me its location.
> Does anyone know more about how locate works who can help me figure this
> out?
>
> Thanks!

Hi -

mlocate, slocate rely on the command 'updatedb' which builds a
database of files that exist.

I don't think this is the proper way of checking, though.... doesn't
the distribution support some sort of package accounting, such as RPM
or deb or something like that?

Other commands which you might find useful are 'which', which will
travers the path and tell you if an executable exists.  'whereis'
which does something similar, or you could simply:

find / -name nginx -perm -o+rx

All of these solutions are pretty dirty.  If I were you I would try to
ensure that first you're using a sane distribution, and second you
trust whoever maintains that installation to not ever, ever install
from a tarball, but to always use the distribution's package
managemnet system.

Thanks
-dant


From nginx-forum at nginx.us  Fri May 13 03:59:15 2011
From: nginx-forum at nginx.us (the_traveller)
Date: Thu, 12 May 2011 19:59:15 -0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <72a81ff3e80ed0b4b78cb81a348826cf@ruby-forum.com>
References: <72a81ff3e80ed0b4b78cb81a348826cf@ruby-forum.com>
Message-ID: <2507a2f6fd2c5e8a743f33ca869dfa82.NginxMailingListEnglish@forum.nginx.org>

Please can you help me to convert this to nginx

<IfModule mod_rewrite.c>
RewriteEngine on

# If you are having problems or are using VirtualDocumentRoot, uncomment
this line and set it to your  directory.
# RewriteBase /maindir/
RewriteOptions MaxRedirects=30
RewriteCond %{ENV:REDIRECT_STATUS} 200
RewriteRule .* - [L]


RewriteCond %{REQUEST_FILENAME} -s [OR]
RewriteCond %{REQUEST_FILENAME} -l [OR]
RewriteCond %{REQUEST_FILENAME} -d

RewriteRule ^.*$ - [NC,L]

RewriteRule ^index(.*)$                                      index.php
[QSA]
RewriteRule ^chat/.*                                         pages.php
[QSA]
RewriteRule ^Songs(.*)$                                     
showcat.php?g=2 [QSA]
RewriteRule ^Jokes(.*)$                                     
showcat.php?g=3 [QSA]
RewriteRule ^Ash3ar(.*)$                                    
showcat.php?g=4 [QSA]
RewriteRule ^iraq/.*                                        showcat.php 
   [QSA]
RewriteRule ^Song/.*                                         sounds.php 
    [QSA]
RewriteRule ^Joke/.*                                         jokes.php  
    [QSA]
RewriteRule ^Poem/.*                                         poems.php  
    [QSA]
RewriteRule ^link/.*                                         advs.php   
    [QSA]
RewriteRule ^search/.*                                       search.php 
    [QSA]
RewriteRule ^adv-conditions(.*)$                                
ad.php?go=conditions      [NC]
RewriteRule ^adv-payments(.*)$                                  
ad.php?go=payments      [NC]

RewriteCond %{REQUEST_FILENAME} -s [OR]
RewriteCond %{REQUEST_FILENAME} -l [OR]
RewriteCond %{REQUEST_FILENAME} -d

RewriteRule ^.*$ - [NC,L]
RewriteRule ^(?:(.*?)(?:/|$))(.*|$)$ $1.php?r=$2 [QSA]


</IfModule>


<IfModule mod_security.c>
   SecFilterEngine Off
   SecFilterScanPOST Off
</IfModule>

<IfModule mod_gzip.c>
mod_gzip_on No
</IfModule>

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,23868,197702#msg-197702



From omega13a at fedtrek.com  Fri May 13 04:19:28 2011
From: omega13a at fedtrek.com (Brandon Amaro)
Date: Thu, 12 May 2011 17:19:28 -0700
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <2507a2f6fd2c5e8a743f33ca869dfa82.NginxMailingListEnglish@forum.nginx.org>
References: <72a81ff3e80ed0b4b78cb81a348826cf@ruby-forum.com>
	<2507a2f6fd2c5e8a743f33ca869dfa82.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4DCC7910.8000800@fedtrek.com>

using http://www.anilcetin.com/convert-apache-htaccess-to-nginx/ , I got 
this:

#ignored: condition 0
#ignored: "-" thing used or unknown variable in regex/rew
if (-e $request_filename){
     set $rule_1 1;
}
if ($request_filename ~ "-l"){
     set $rule_1 1;
}
if (-d $request_filename){
     set $rule_1 1;
}
if ($rule_1 = "1"){
#ignored: "-" thing used or unknown variable in regex/rew
}
     rewrite ^/index(.*)$ /;
     rewrite ^/chat/.* /;
     rewrite ^/Songs(.*)$ /;
     rewrite ^/Jokes(.*)$ /;
     rewrite ^/Ash3ar(.*)$ /;
     rewrite ^/iraq/.* /;
     rewrite ^/Song/.* /;
     rewrite ^/Joke/.* /;
     rewrite ^/Poem/.* /;
     rewrite ^/link/.* /;
     rewrite ^/search/.* /;
     rewrite ^/adv-conditions(.*)$ /;
     rewrite ^/adv-payments(.*)$ /;
if (-e $request_filename){
     set $rule_15 1;
}
if ($request_filename ~ "-l"){
     set $rule_15 1;
}
if (-d $request_filename){
     set $rule_15 1;
}
if ($rule_15 = "1"){
#ignored: "-" thing used or unknown variable in regex/rew
}
     rewrite ^/(?:(.*?)(?:/|$))(.*|$)$ /$1.php?r=$2;

-- 
omega13a
Owner and Founder of UFT
http://www.fedtrek.com


On 05/12/2011 04:59 PM, the_traveller wrote:
> <IfModule mod_rewrite.c>
> RewriteEngine on
>
> # If you are having problems or are using VirtualDocumentRoot, uncomment
> this line and set it to your  directory.
> # RewriteBase/maindir/
> RewriteOptions MaxRedirects=30
> RewriteCond %{ENV:REDIRECT_STATUS} 200
> RewriteRule .* - [L]
>
>
> RewriteCond %{REQUEST_FILENAME} -s [OR]
> RewriteCond %{REQUEST_FILENAME} -l [OR]
> RewriteCond %{REQUEST_FILENAME} -d
>
> RewriteRule ^.*$ - [NC,L]
>
> RewriteRule ^index(.*)$                                      index.php
> [QSA]
> RewriteRule ^chat/.*                                         pages.php
> [QSA]
> RewriteRule ^Songs(.*)$
> showcat.php?g=2 [QSA]
> RewriteRule ^Jokes(.*)$
> showcat.php?g=3 [QSA]
> RewriteRule ^Ash3ar(.*)$
> showcat.php?g=4 [QSA]
> RewriteRule ^iraq/.*                                        showcat.php
>     [QSA]
> RewriteRule ^Song/.*                                         sounds.php
>      [QSA]
> RewriteRule ^Joke/.*                                         jokes.php
>      [QSA]
> RewriteRule ^Poem/.*                                         poems.php
>      [QSA]
> RewriteRule ^link/.*                                         advs.php
>      [QSA]
> RewriteRule ^search/.*                                       search.php
>      [QSA]
> RewriteRule ^adv-conditions(.*)$
> ad.php?go=conditions      [NC]
> RewriteRule ^adv-payments(.*)$
> ad.php?go=payments      [NC]
>
> RewriteCond %{REQUEST_FILENAME} -s [OR]
> RewriteCond %{REQUEST_FILENAME} -l [OR]
> RewriteCond %{REQUEST_FILENAME} -d
>
> RewriteRule ^.*$ - [NC,L]
> RewriteRule ^(?:(.*?)(?:/|$))(.*|$)$ $1.php?r=$2 [QSA]
>
>
> </IfModule>
>
>
> <IfModule mod_security.c>
>     SecFilterEngine Off
>     SecFilterScanPOST Off
> </IfModule>
>
> <IfModule mod_gzip.c>
> mod_gzip_on No
> </IfModule>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110512/51a7e00e/attachment-0001.html>

From nginx-forum at nginx.us  Fri May 13 05:08:36 2011
From: nginx-forum at nginx.us (the_traveller)
Date: Thu, 12 May 2011 21:08:36 -0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <2507a2f6fd2c5e8a743f33ca869dfa82.NginxMailingListEnglish@forum.nginx.org>
References: <72a81ff3e80ed0b4b78cb81a348826cf@ruby-forum.com>
	<2507a2f6fd2c5e8a743f33ca869dfa82.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20303b9b6853ad3803b0460878e26cce.NginxMailingListEnglish@forum.nginx.org>

well first of all thank you for you quick reply ..
but i've tried this tool and it's not working...
i there any way to know the basics of the ngnix???

other wise look at the rewrite model 

rewrite ^/index(.*)$ /;
rewrite ^/chat/.* /;
rewrite ^/Songs(.*)$ /;
rewrite ^/Jokes(.*)$ /;
rewrite ^/Ash3ar(.*)$ /;
rewrite ^/iraq/.* /;
rewrite ^/Song/.* /;
rewrite ^/Joke/.* /;
rewrite ^/Poem/.* /;
rewrite ^/link/.* /;
rewrite ^/search/.* /;
rewrite ^/adv-conditions(.*)$ /;
rewrite ^/adv-payments(.*)$ /;


it haven't consider the file file names ... so i'm little bit confused
about that tool
and so it didn't work

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,23868,197717#msg-197717



From bbskill at gmail.com  Fri May 13 06:58:54 2011
From: bbskill at gmail.com (Liu bbskill)
Date: Fri, 13 May 2011 10:58:54 +0800
Subject: Why doesn't Nginx support fastcgi keepalive connection?
Message-ID: <BANLkTim08c0w5y-Mm=MdEVpDOcXcGqsspQ@mail.gmail.com>

Hi, all

    It seems that nginx will communicate with fastcgi using short
connection, and with heavy load, nginx and fastcgi will cause large number
of TIME_WAIT.  It's annoying.

   I am wondering why nginx doesn't support fastcgi keepalive connection(*
FCGI_KEEP_CONN*)?

   Are there some special considerations?


-- 
JinBiao Liu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110513/36e6a4b1/attachment.html>

From nginx-forum at nginx.us  Fri May 13 08:49:53 2011
From: nginx-forum at nginx.us (spacereactor)
Date: Fri, 13 May 2011 00:49:53 -0400
Subject: After upgrading nginx from 0.8.54 to 1.0.2,
	getting blank page with drupal context module install
Message-ID: <c9a100f6f3789b9854a169296c764e1e.NginxMailingListEnglish@forum.nginx.org>

This is the error log from server 
2011/05/11 17:40:02 [error] 3358#0: *1 FastCGI sent in stderr: "PHP
Fatal error:  require() [<a
href='function.require'>function.require</a>]: Failed opening required
'context.core.inc' (include_path='.:/usr/local/lib/php') in ???
 on
line 3" while reading response header from upstream, client:
220.255.1.81, server: mydomain.com, request: "GET / HTTP/1.1", upstream:
"fastcgi://unix:/var/run/php5-fpm.sock:", host: "mydomain.com"
2011/05/11 17:40:05 [error] 3358#0: *1 FastCGI sent in stderr: "PHP
Fatal error:  require() [<a
href='function.require'>function.require</a>]: Failed opening required
'context.core.inc' (include_path='.:/usr/local/lib/php') in x
?
 on
line 3" while reading response header from upstream, client:
220.255.1.81, server: mydomain.com, request: "GET / HTTP/1.1", upstream:
"fastcgi://unix:/var/run/php5-fpm.sock:", host: "mydomain.com"
2011/05/11 17:45:08 [error] 3383#0: *1 no user/password was provided for
basic authentication, client: 220.255.1.85, server: mydomain.com,
request: "GET /install.php HTTP/1.1", host: "mydomain.com"
2011/05/11 17:50:22 [error] 3383#0: *115 FastCGI sent in stderr: "PHP
Fatal error:  require() [<a
href='function.require'>function.require</a>]: Failed opening required
'context.core.inc' (include_path='.:/usr/local/lib/php') in ???
 on
line 3" while reading response header from upstream, client:
220.255.1.52, server: mydomain.com, request: "POST
/admin/build/modules/list/confirm HTTP/1.1", upstream:
"fastcgi://unix:/var/run/php5-fpm.sock:", host: "mydomain.com",
referrer: "http://mydomain.com/admin/build/modules/list/confirm"

My using linode server, nginx 1.0.2 / php 5.2.17 / php-fpm  / mariadb
5.2
If i purge nginx 1.0.2 and install back 0.8.54 the drupal context work.
I also try with a fresh install of drupal and only install context
modules and i getting error with nginx 1.0.2

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197733,197733#msg-197733



From appa at perusio.net  Fri May 13 09:20:41 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Fri, 13 May 2011 06:20:41 +0100
Subject: After upgrading nginx from 0.8.54 to 1.0.2,
	getting blank page with drupal context module install
In-Reply-To: <c9a100f6f3789b9854a169296c764e1e.NginxMailingListEnglish@forum.nginx.org>
References: <c9a100f6f3789b9854a169296c764e1e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87tycz9nx2.wl%appa@perusio.net>

On 13 Mai 2011 05h49 WEST, nginx-forum at nginx.us wrote:

That seems to be a PHP problem. 

Try to restart fpm and see if the problem remains. 
Are you using APC?


> This is the error log from server 2011/05/11 17:40:02 [error]
> 3358#0: *1 FastCGI sent in stderr: "PHP Fatal error: require() [<a
> href='function.require'>function.require</a>]: Failed opening
> required 'context.core.inc' (include_path='.:/usr/local/lib/php') in
> ???
 on line 3" while reading response header from upstream,
> client: 220.255.1.81, server: mydomain.com, request: "GET /
> HTTP/1.1", upstream: "fastcgi://unix:/var/run/php5-fpm.sock:", host:
> "mydomain.com" 2011/05/11 17:40:05 [error] 3358#0: *1 FastCGI sent
> in stderr: "PHP Fatal error: require() [<a
> href='function.require'>function.require</a>]: Failed opening
> required 'context.core.inc' (include_path='.:/usr/local/lib/php') in
> x
?
 on line 3" while reading response header from upstream,
> client: 220.255.1.81, server: mydomain.com, request: "GET /
> HTTP/1.1", upstream: "fastcgi://unix:/var/run/php5-fpm.sock:", host:
> "mydomain.com" 2011/05/11 17:45:08 [error] 3383#0: *1 no
> user/password was provided for basic authentication, client:
> 220.255.1.85, server: mydomain.com, request: "GET /install.php
> HTTP/1.1", host: "mydomain.com" 2011/05/11 17:50:22 [error] 3383#0:
> *115 FastCGI sent in stderr: "PHP Fatal error: require() [<a
> href='function.require'>function.require</a>]: Failed opening
> required 'context.core.inc' (include_path='.:/usr/local/lib/php') in
> ???
 on line 3" while reading response header from upstream,
> client: 220.255.1.52, server: mydomain.com, request: "POST
> /admin/build/modules/list/confirm HTTP/1.1", upstream:
> "fastcgi://unix:/var/run/php5-fpm.sock:", host: "mydomain.com",
> referrer: "http://mydomain.com/admin/build/modules/list/confirm"

--- appa



From nginx-forum at nginx.us  Fri May 13 09:28:25 2011
From: nginx-forum at nginx.us (spacereactor)
Date: Fri, 13 May 2011 01:28:25 -0400
Subject: After upgrading nginx from 0.8.54 to 1.0.2,
	getting blank page with drupal context module install
In-Reply-To: <c9a100f6f3789b9854a169296c764e1e.NginxMailingListEnglish@forum.nginx.org>
References: <c9a100f6f3789b9854a169296c764e1e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4ef952b878dd0ca524e672548548de08.NginxMailingListEnglish@forum.nginx.org>

try restart php-fpm and also test with APC-3.1.7 but result is the same
error, i even try update the APC to APC-3.18 but also the same problem.
If it the context module php issue, than should nginx 0.8.54 also have
that problem but drupal context work under nginx 0.8.54 but not nginx
1.0.2

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197733,197740#msg-197740



From appa at perusio.net  Fri May 13 16:21:15 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Fri, 13 May 2011 13:21:15 +0100
Subject: After upgrading nginx from 0.8.54 to 1.0.2,
	getting blank page with drupal context module install
In-Reply-To: <4ef952b878dd0ca524e672548548de08.NginxMailingListEnglish@forum.nginx.org>
References: <c9a100f6f3789b9854a169296c764e1e.NginxMailingListEnglish@forum.nginx.org>
	<4ef952b878dd0ca524e672548548de08.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87sjsiaj0k.wl%appa@perusio.net>

On 13 Mai 2011 06h28 WEST, nginx-forum at nginx.us wrote:

> try restart php-fpm and also test with APC-3.1.7 but result is the
> same error, i even try update the APC to APC-3.18 but also the same
> problem.  If it the context module php issue, than should nginx
> 0.8.54 also have that problem but drupal context work under nginx
> 0.8.54 but not nginx 1.0.2

I propose we move this conversation to the g.d.o nginx group
http://groups.drupal.org/nginx

Context works ok for me with 1.0.2.

--- appa



From lists at ruby-forum.com  Fri May 13 17:02:28 2011
From: lists at ruby-forum.com (John Travolota)
Date: Fri, 13 May 2011 15:02:28 +0200
Subject: kernel: nginx[18233]: segfault at 0000000000000008 rip
	000000000043edf8 rsp 00007fff34a21fa0 error 4
Message-ID: <d6ba00b08aeeb18ecc9a0873cdd53e5e@ruby-forum.com>

I having this message which i posted in subject repeating a lot in
/var/log/messages,is this some kind of bug?Nginx ver 1.0.1

-- 
Posted via http://www.ruby-forum.com/.


From mdounin at mdounin.ru  Fri May 13 17:09:31 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 13 May 2011 17:09:31 +0400
Subject: Why doesn't Nginx support fastcgi keepalive connection?
In-Reply-To: <BANLkTim08c0w5y-Mm=MdEVpDOcXcGqsspQ@mail.gmail.com>
References: <BANLkTim08c0w5y-Mm=MdEVpDOcXcGqsspQ@mail.gmail.com>
Message-ID: <20110513130931.GW42265@mdounin.ru>

Hello!

On Fri, May 13, 2011 at 10:58:54AM +0800, Liu bbskill wrote:

> Hi, all
> 
>     It seems that nginx will communicate with fastcgi using short
> connection, and with heavy load, nginx and fastcgi will cause large number
> of TIME_WAIT.  It's annoying.
> 
>    I am wondering why nginx doesn't support fastcgi keepalive connection(*
> FCGI_KEEP_CONN*)?
> 
>    Are there some special considerations?

There are some experimental patches available which add fastcgi 
keepalive support (with upstream keepalive module).

I'll post refreshed version in a few weeks.

Maxim Dounin


From mdounin at mdounin.ru  Fri May 13 17:18:56 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 13 May 2011 17:18:56 +0400
Subject: kernel: nginx[18233]: segfault at 0000000000000008 rip
	000000000043edf8 rsp 00007fff34a21fa0 error 4
In-Reply-To: <d6ba00b08aeeb18ecc9a0873cdd53e5e@ruby-forum.com>
References: <d6ba00b08aeeb18ecc9a0873cdd53e5e@ruby-forum.com>
Message-ID: <20110513131856.GX42265@mdounin.ru>

Hello!

On Fri, May 13, 2011 at 03:02:28PM +0200, John Travolota wrote:

> I having this message which i posted in subject repeating a lot in
> /var/log/messages,is this some kind of bug?Nginx ver 1.0.1

Segmentation fault is always some kind of bug (though not always 
in the programm which actually segfaults).

Please provide at least nginx -V output, config and backtrace.  
See here for more details:

http://wiki.nginx.org/Debugging

Maxim Dounin


From lists at ruby-forum.com  Fri May 13 17:23:04 2011
From: lists at ruby-forum.com (John Travolota)
Date: Fri, 13 May 2011 15:23:04 +0200
Subject: kernel: nginx[18233]: segfault at 0000000000000008 rip
	000000000043edf8 rsp 00007fff34a21fa0 err
In-Reply-To: <d6ba00b08aeeb18ecc9a0873cdd53e5e@ruby-forum.com>
References: <d6ba00b08aeeb18ecc9a0873cdd53e5e@ruby-forum.com>
Message-ID: <fb695559c57c62f97804f76d1e758cf6@ruby-forum.com>

Here is output of V along with full conf options,not sure how to do 
backtrace.
nginx: nginx version: nginx/1.0.1
nginx: built by gcc 4.1.2 20080704 (Red Hat 4.1.2-48)
nginx: TLS SNI support disabled 
nginx: configure arguments: --sbin-path=/usr/local/sbin 
--with-http_ssl_module --with-http_geoip_module 
--without-http_autoindex_module --without-mail_pop3_module 
--without-mail_imap_module --without-mail_smtp_module 
--with-http_stub_st
atus_module

-- 
Posted via http://www.ruby-forum.com/.


From b.yordanov at exsisto.com  Fri May 13 17:31:06 2011
From: b.yordanov at exsisto.com (Boyko Yordanov)
Date: Fri, 13 May 2011 16:31:06 +0300
Subject: kernel: nginx[18233]: segfault at 0000000000000008 rip
	000000000043edf8 rsp 00007fff34a21fa0 err
In-Reply-To: <fb695559c57c62f97804f76d1e758cf6@ruby-forum.com>
References: <d6ba00b08aeeb18ecc9a0873cdd53e5e@ruby-forum.com>
	<fb695559c57c62f97804f76d1e758cf6@ruby-forum.com>
Message-ID: <86C89487-2A95-4CEE-86D2-CC6994880AA6@exsisto.com>

He gave you a hint: http://wiki.nginx.org/Debugging

On May 13, 2011, at 4:23 PM, John Travolota wrote:

> not sure how to do 
> backtrace.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110513/20858a19/attachment.html>

From mp3geek at gmail.com  Fri May 13 19:02:52 2011
From: mp3geek at gmail.com (Ryan B)
Date: Sat, 14 May 2011 03:02:52 +1200
Subject: rewrite n00b
In-Reply-To: <20110512052052.GC99505@sysoev.ru>
References: <BANLkTin7mX3yD8r1Uwq+mwaW-SMaBCWwYw@mail.gmail.com>
	<20110512052052.GC99505@sysoev.ru>
Message-ID: <BANLkTinZNGM40WNEL847zx++_iQ6sn7mzg@mail.gmail.com>

Awesome thanks :)

I've got this, to redirect the requests, but If I request a file from
the orginal  host, it won't redirect to the same file on the ip.

rewrite ^ http://113.214.191.192 permanent;

So http://domain.com/files.txt -> redirects to
http://113.214.191.192/files.txt (without the need to specifiy each
file name)

Thanks for the help



On Thu, May 12, 2011 at 5:20 PM, Igor Sysoev <igor at sysoev.ru> wrote:
> On Thu, May 12, 2011 at 02:13:03PM +1200, Ryan B wrote:
>> Hi,
>> Just wanting to enabled a rewrite of the following url,
>>
>> Redirecting all requests from http://www.domain.com/site/forums and
>> http://www.domain.com/forums to http://forums.domain.com
>
> server {
> ? ?server_name ?www.domain.com;
>
> ? ?location /forums {
> ? ? ? ?rewrite ?^/forums(.*)$ ?http://forums.domain.com$1;
> ? ?}
>
> ? ?location /site/forums {
> ? ? ? ?rewrite ?^/site/forums(.*)$ ?http://forums.domain.com$1;
> ? ?}
>
> ? ?...
>
>
> --
> Igor Sysoev
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>


From igor at sysoev.ru  Fri May 13 22:16:51 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 13 May 2011 22:16:51 +0400
Subject: rewrite n00b
In-Reply-To: <BANLkTinZNGM40WNEL847zx++_iQ6sn7mzg@mail.gmail.com>
References: <BANLkTin7mX3yD8r1Uwq+mwaW-SMaBCWwYw@mail.gmail.com>
	<20110512052052.GC99505@sysoev.ru>
	<BANLkTinZNGM40WNEL847zx++_iQ6sn7mzg@mail.gmail.com>
Message-ID: <20110513181651.GC91483@sysoev.ru>

On Sat, May 14, 2011 at 03:02:52AM +1200, Ryan B wrote:
> Awesome thanks :)
> 
> I've got this, to redirect the requests, but If I request a file from
> the orginal  host, it won't redirect to the same file on the ip.
> 
> rewrite ^ http://113.214.191.192 permanent;
> 
> So http://domain.com/files.txt -> redirects to
> http://113.214.191.192/files.txt (without the need to specifiy each
> file name)

   rewrite ^ http://113.214.191.192$request_uri?  permanent;


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May 13 22:54:06 2011
From: nginx-forum at nginx.us (holstebm)
Date: Fri, 13 May 2011 14:54:06 -0400
Subject: IPv6 Issues / Question
Message-ID: <170e85c7ee18ce77bcc3ba9560b19f84.NginxMailingListEnglish@forum.nginx.org>

Last night I had my provider (Linode) add IPv6 support for my server.  I
also had them assign me an IPv6 pool to use.

I have Nginx 1.0.2 compiled with IPv6 support, restarted my Nginx and
everything was fine all night and most of the morning.  However, within
minutes of me adding AAAA records to my domains, Nginx stopped and
running and took the IPv4 networking down with it.  I hadn't even
restarted Nginx after adding the records.

Upon reboot with Nginx disabled, everything runs fine.  I start Nginx
and make a single IPv4 page request, the network is completely gone
again  on all ports.

I don't believe in coincidences so what could adding AAAA records to
domains Nginx is listening for on port 80 trigger?  Nginx had been
running without problems for 12 hours with IPv6 enabled on the server
with an assigned global scope IPv6 address.  Within minutes of me adding
those AAAA records, the trouble began.

I don't see anything in the logs that helps.  Here is my configuration
for Nginx:

nginx: configure arguments: 
--sbin-path=/usr/sbin/nginx
 --conf-path=/etc/nginx/nginx.conf
 --user=nginx
 --group=nginx
 --http-client-body-temp-path=/var/lib/nginx/tmp/client_body
 --http-proxy-temp-path=/var/lib/nginx/tmp/proxy
 --http-fastcgi-temp-path=/var/lib/nginx/tmp/fastcgi
 --error-log-path=/var/log/nginx/error.log
 --http-log-path=/var/log/nginx/access.log
 --prefix=/usr/share/nginx
 --pid-path=/var/run/nginx.pid
 --add-module=../headers-more-nginx-module
 --add-module=../ngx_http_redis-0.3.1
 --add-module=../ngx_http_upstream_keepalive
 --add-module=../memc-nginx-module
 --add-module=../ngx_cache_purge
 --add-module=/usr/local/rvm/gems/ruby-1.9.2-p0 at rails3/gems/passenger-3.0.2/ext/nginx
 --add-module=../ngx_http_log_request_speed
 --add-module=../mod_strip
 --with-ipv6
 --with-http_ssl_module
 --with-http_gzip_static_module
 --with-http_xslt_module
 --with-http_image_filter_module
 --with-http_geoip_module
 --with-http_sub_module
 --with-http_realip_module
 --with-http_stub_status_module
 --with-http_flv_module
 --without-mail_pop3_module
 --without-mail_imap_module
 --without-mail_smtp_module
 --without-select_module
 --without-poll_module
 --with-openssl=../openssl-1.0.0d
 --with-pcre=../pcre-8.12
 --with-zlib=../zlib-1.2.5
 --with-cpu-opt=Xeon
 --with-cc-opt='-pipe -s -O3 -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions
-fstack-protector --param=ssp-buffer-size=4 -mmmx -msse -msse2 -msse3
-mfpmath=sse -mtune=native -march=native -enable-new-dtags -DNODEBUG
-DNDEBUG -enable-stdcall-fixup -frerun-loop-opt -ftracer'

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197875,197875#msg-197875



From nginx-forum at nginx.us  Fri May 13 23:05:22 2011
From: nginx-forum at nginx.us (holstebm)
Date: Fri, 13 May 2011 15:05:22 -0400
Subject: IPv6 Issues / Question
In-Reply-To: <170e85c7ee18ce77bcc3ba9560b19f84.NginxMailingListEnglish@forum.nginx.org>
References: <170e85c7ee18ce77bcc3ba9560b19f84.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <775d45c0b90e18499e713b91724261d0.NginxMailingListEnglish@forum.nginx.org>

I should add that disabling IPv6 this way:

sysctl -w net.ipv6.conf.all.disable_ipv6=1
sysctl -w net.ipv6.conf.default.disable_ipv6=1
sysctl -w net.ipv6.conf.lo.disable_ipv6=1

without making any changes to the Nginx configuration, works.  So it is
definitely related to IPv6 and Nginx.

TIA!

Bret

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197875,197877#msg-197877



From igor at sysoev.ru  Fri May 13 23:13:45 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 13 May 2011 23:13:45 +0400
Subject: IPv6 Issues / Question
In-Reply-To: <170e85c7ee18ce77bcc3ba9560b19f84.NginxMailingListEnglish@forum.nginx.org>
References: <170e85c7ee18ce77bcc3ba9560b19f84.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110513191345.GF91483@sysoev.ru>

On Fri, May 13, 2011 at 02:54:06PM -0400, holstebm wrote:
> Last night I had my provider (Linode) add IPv6 support for my server.  I
> also had them assign me an IPv6 pool to use.
> 
> I have Nginx 1.0.2 compiled with IPv6 support, restarted my Nginx and
> everything was fine all night and most of the morning.  However, within
> minutes of me adding AAAA records to my domains, Nginx stopped and
> running and took the IPv4 networking down with it.  I hadn't even
> restarted Nginx after adding the records.
> 
> Upon reboot with Nginx disabled, everything runs fine.  I start Nginx
> and make a single IPv4 page request, the network is completely gone
> again  on all ports.

What do you mean by "took the IPv4 networking down" ? Does it mean
that any (inlcuding ssh, etc) IPv4 traffic goes down, or only HTTP or what ?


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May 13 23:25:47 2011
From: nginx-forum at nginx.us (holstebm)
Date: Fri, 13 May 2011 15:25:47 -0400
Subject: IPv6 Issues / Question
In-Reply-To: <20110513191345.GF91483@sysoev.ru>
References: <20110513191345.GF91483@sysoev.ru>
Message-ID: <d6cc22796f6a0c5d048a2a1b626d3dc7.NginxMailingListEnglish@forum.nginx.org>

Literally everything.  SSH - everything.  This is a VPS (Linode) with a
Lish AJAX console that normally works even when all other networking is
disabled and even it was down.  I was literally blind and the only
option left was to reboot each time.  It took several reboots to narrow
it down to a particular service.

Please let me know if there is a particular log I should look at.  I've
looked at the messages and syslog as well as the access and error logs
for nginx and they are all missing entries during the time networking
was down.

Thanks!  I've been using Nginx for years barely a hiccup until this.

Bret

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197875,197883#msg-197883



From nginx-forum at nginx.us  Fri May 13 23:28:39 2011
From: nginx-forum at nginx.us (holstebm)
Date: Fri, 13 May 2011 15:28:39 -0400
Subject: IPv6 Issues / Question
In-Reply-To: <d6cc22796f6a0c5d048a2a1b626d3dc7.NginxMailingListEnglish@forum.nginx.org>
References: <20110513191345.GF91483@sysoev.ru>
	<d6cc22796f6a0c5d048a2a1b626d3dc7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <980b1fdb965f8d91fd342cdd1785f3fd.NginxMailingListEnglish@forum.nginx.org>

I haven't eliminated the possibility that Nginx is the symptom and not
the cause.  Of course, under the circumstances, I'm not sure how I might
do that.  But I am willing to provide anything you need to help get to
the bottom of this.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197875,197885#msg-197885



From igor at sysoev.ru  Fri May 13 23:36:57 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 13 May 2011 23:36:57 +0400
Subject: IPv6 Issues / Question
In-Reply-To: <d6cc22796f6a0c5d048a2a1b626d3dc7.NginxMailingListEnglish@forum.nginx.org>
References: <20110513191345.GF91483@sysoev.ru>
	<d6cc22796f6a0c5d048a2a1b626d3dc7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110513193657.GG91483@sysoev.ru>

On Fri, May 13, 2011 at 03:25:47PM -0400, holstebm wrote:
> Literally everything.  SSH - everything.  This is a VPS (Linode) with a
> Lish AJAX console that normally works even when all other networking is
> disabled and even it was down.  I was literally blind and the only
> option left was to reboot each time.  It took several reboots to narrow
> it down to a particular service.

I believe this is a bug in Linux or VPS.
A user mode program even run with root privilege that makes only usual
listen/bind/setsockopt/accept/etc syscalls on IPv6 sockets should not
cause such issues.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May 13 23:47:29 2011
From: nginx-forum at nginx.us (holstebm)
Date: Fri, 13 May 2011 15:47:29 -0400
Subject: IPv6 Issues / Question
In-Reply-To: <20110513193657.GG91483@sysoev.ru>
References: <20110513193657.GG91483@sysoev.ru>
Message-ID: <463cd14d8c152c338ea9fe5612f22c9d.NginxMailingListEnglish@forum.nginx.org>

Okay, thanks, I'll into that bit more.  This is a production machine
which, right now, limits me a bit.  I is running Ubuntu 10.04 with one
of the kernels from Linode.  Sometime overnight this weekend, I'll play
with it a bit more.  I can use other kernels or compile my own, if
necessary.

I was hoping what you thought about the AAAA records being added and the
appearance of the issue might be helpful as it seems strange.  Perhaps a
routing problem of some sort when Nginx sent out a reply?

Thanks for you help!

Bret

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197875,197892#msg-197892



From igor at sysoev.ru  Fri May 13 23:49:42 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 13 May 2011 23:49:42 +0400
Subject: IPv6 Issues / Question
In-Reply-To: <463cd14d8c152c338ea9fe5612f22c9d.NginxMailingListEnglish@forum.nginx.org>
References: <20110513193657.GG91483@sysoev.ru>
	<463cd14d8c152c338ea9fe5612f22c9d.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110513194942.GH91483@sysoev.ru>

On Fri, May 13, 2011 at 03:47:29PM -0400, holstebm wrote:
> Okay, thanks, I'll into that bit more.  This is a production machine
> which, right now, limits me a bit.  I is running Ubuntu 10.04 with one
> of the kernels from Linode.  Sometime overnight this weekend, I'll play
> with it a bit more.  I can use other kernels or compile my own, if
> necessary.
> 
> I was hoping what you thought about the AAAA records being added and the
> appearance of the issue might be helpful as it seems strange.  Perhaps a
> routing problem of some sort when Nginx sent out a reply?

AAAA record may just cause IPv6 traffic to the host.
And the traffic may cause kernel issues.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Sat May 14 00:01:09 2011
From: nginx-forum at nginx.us (jfix)
Date: Fri, 13 May 2011 16:01:09 -0400
Subject: trouble with forward proxying
In-Reply-To: <4e5c2db6e47b5d013dfbb23fa21a4b8a.NginxMailingListEnglish@forum.nginx.org>
References: <4e5c2db6e47b5d013dfbb23fa21a4b8a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <10bb050b819b0ef52aebb1a5f4b0890e.NginxMailingListEnglish@forum.nginx.org>

thanks, I have switched to squid. Jakob.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196115,197896#msg-197896



From jeffrpang at gmail.com  Sat May 14 00:06:47 2011
From: jeffrpang at gmail.com (Jeff Pang)
Date: Sat, 14 May 2011 04:06:47 +0800
Subject: trouble with forward proxying
In-Reply-To: <10bb050b819b0ef52aebb1a5f4b0890e.NginxMailingListEnglish@forum.nginx.org>
References: <4e5c2db6e47b5d013dfbb23fa21a4b8a.NginxMailingListEnglish@forum.nginx.org>
	<10bb050b819b0ef52aebb1a5f4b0890e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTikWjcajRXYW-V=SpGC+Zk25cXXcdg@mail.gmail.com>

2011/5/14 jfix <nginx-forum at nginx.us>:
> thanks, I have switched to squid. Jakob.
>

Squid is Proxy + Caching, it will improve the speed if you have lots
of static files.
But Squid is not compatible as well as Nginx for HTTP/1.1.

-- 
Jeff Pang
www.DNSbed.com


From nginx-forum at nginx.us  Sat May 14 00:09:09 2011
From: nginx-forum at nginx.us (jfix)
Date: Fri, 13 May 2011 16:09:09 -0400
Subject: gzip compression not taken into account?
Message-ID: <a1bb7953bf979c5dba6b2802f7e5b67a.NginxMailingListEnglish@forum.nginx.org>

Hi, I have enabled gzip compression following the wiki instructions like
this:

location / {
 #....

 gzip on;
 gzip_min_length 1000;
 gzip_proxied expired no-cache no-store private auth;
 gzip_types text/plain application/xml text/javascript text/css
application/x-javascript;
 gzip_disable "MSIE [1-6]\.";
 
 #...
}

unfortunately, both Yahoo's yslow and Google's pagespeed tell me that I
should enable compression for better performance. 

How can I find out why compression doesn't seem to happen?  I've
selected the mime-types after having looked at them in chrome's web
inspector.

Thank you in advance,
Jakob.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197898,197898#msg-197898



From dan.trainor at gmail.com  Sat May 14 00:19:55 2011
From: dan.trainor at gmail.com (Dan Trainor)
Date: Fri, 13 May 2011 13:19:55 -0700
Subject: gzip compression not taken into account?
In-Reply-To: <a1bb7953bf979c5dba6b2802f7e5b67a.NginxMailingListEnglish@forum.nginx.org>
References: <a1bb7953bf979c5dba6b2802f7e5b67a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTin3nTOynPUB694WHeC2jzSk7UxV1Q@mail.gmail.com>

Hi -

Are those tools misreporting?  What does the content-encoding: header say?
On May 13, 2011 1:09 PM, "jfix" <nginx-forum at nginx.us> wrote:
> Hi, I have enabled gzip compression following the wiki instructions like
> this:
>
> location / {
> #....
>
> gzip on;
> gzip_min_length 1000;
> gzip_proxied expired no-cache no-store private auth;
> gzip_types text/plain application/xml text/javascript text/css
> application/x-javascript;
> gzip_disable "MSIE [1-6]\.";
>
> #...
> }
>
> unfortunately, both Yahoo's yslow and Google's pagespeed tell me that I
> should enable compression for better performance.
>
> How can I find out why compression doesn't seem to happen? I've
> selected the mime-types after having looked at them in chrome's web
> inspector.
>
> Thank you in advance,
> Jakob.
>
> Posted at Nginx Forum:
http://forum.nginx.org/read.php?2,197898,197898#msg-197898
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110513/7b63c990/attachment.html>

From jocelyn.mocquant at gmail.com  Sat May 14 03:00:27 2011
From: jocelyn.mocquant at gmail.com (Jocelyn)
Date: Fri, 13 May 2011 16:00:27 -0700
Subject: gzip compression not taken into account?
In-Reply-To: <BANLkTin3nTOynPUB694WHeC2jzSk7UxV1Q@mail.gmail.com>
References: <a1bb7953bf979c5dba6b2802f7e5b67a.NginxMailingListEnglish@forum.nginx.org>
	<BANLkTin3nTOynPUB694WHeC2jzSk7UxV1Q@mail.gmail.com>
Message-ID: <BANLkTimm6m44MPGSPeE-QQjxRapkjwJGQw@mail.gmail.com>

Hi,

Maybe you'are encountering what Steve Souders call the "Turtle Tapping".

You need to check the Accept-Encoding HTTP header from the browser side
(need to be: Accept-Encoding: gzip, deflate). In some cases (in order to
analyse faster the content of the traffic), web-proxies and Antivirus are
mangled the Accept-Encoding HTTP header.

--
Regards
       Jocelyn Mocquant


On Fri, May 13, 2011 at 1:19 PM, Dan Trainor <dan.trainor at gmail.com> wrote:

> Hi -
>
> Are those tools misreporting?  What does the content-encoding: header say?
> On May 13, 2011 1:09 PM, "jfix" <nginx-forum at nginx.us> wrote:
> > Hi, I have enabled gzip compression following the wiki instructions like
> > this:
> >
> > location / {
> > #....
> >
> > gzip on;
> > gzip_min_length 1000;
> > gzip_proxied expired no-cache no-store private auth;
> > gzip_types text/plain application/xml text/javascript text/css
> > application/x-javascript;
> > gzip_disable "MSIE [1-6]\.";
> >
> > #...
> > }
> >
> > unfortunately, both Yahoo's yslow and Google's pagespeed tell me that I
> > should enable compression for better performance.
> >
> > How can I find out why compression doesn't seem to happen? I've
> > selected the mime-types after having looked at them in chrome's web
> > inspector.
> >
> > Thank you in advance,
> > Jakob.
> >
> > Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,197898,197898#msg-197898
> >
> >
> > _______________________________________________
> > nginx mailing list
> > nginx at nginx.org
> > http://nginx.org/mailman/listinfo/nginx
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110513/c324747d/attachment-0001.html>

From nginx-forum at nginx.us  Sat May 14 03:24:12 2011
From: nginx-forum at nginx.us (Jocelyn Mocquant)
Date: Fri, 13 May 2011 19:24:12 -0400
Subject: Keepalived Connections Reset after reloading the configuration (HUP
	Signal)
Message-ID: <46781f68e473c0d6a12e4a3e9788a38e.NginxMailingListEnglish@forum.nginx.org>

Hello,

   I'm dealing with a problem. When reloading the nginx configuration,
all keepalived connections receive the  TCP reset flag after I send a
HUP signal to the master process. If I comment the line responsible for
enabling the keepalive feature in the configuration, the problem
disappear (nginx version is 0.9.7). 


Thanks in advance,
            Jocelyn Mocquant

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197927,197927#msg-197927



From nginx-forum at nginx.us  Sat May 14 13:28:19 2011
From: nginx-forum at nginx.us (spacereactor)
Date: Sat, 14 May 2011 05:28:19 -0400
Subject: After upgrading nginx from 0.8.54 to 1.0.2,
	getting blank page with drupal context module install
In-Reply-To: <87sjsiaj0k.wl%appa@perusio.net>
References: <87sjsiaj0k.wl%appa@perusio.net>
Message-ID: <6791015a56febd3a29bec29a25ea3fc0.NginxMailingListEnglish@forum.nginx.org>

is indeed APC-3.18 that cause the problem, i re-compile the php and use
APC-3.17 it work, and purge all and re-compile php and APC-3.18 and it
doesn't work.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197733,197970#msg-197970



From nginx-forum at nginx.us  Sat May 14 16:19:45 2011
From: nginx-forum at nginx.us (jeanjan)
Date: Sat, 14 May 2011 08:19:45 -0400
Subject: cgi files in plain text
In-Reply-To: <875369c24424bd6f71a4658c85bce47b.NginxMailingListEnglish@forum.nginx.org>
References: <875369c24424bd6f71a4658c85bce47b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <beb82f65a5677d4ad4c4ed785a4697bf.NginxMailingListEnglish@forum.nginx.org>

Thank you, I found out how to access collectd-web cause it has its own
webserver.
:s

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197100,197988#msg-197988



From nginx-forum at nginx.us  Sat May 14 19:16:22 2011
From: nginx-forum at nginx.us (inph)
Date: Sat, 14 May 2011 11:16:22 -0400
Subject: nginx 1.0.2 intermittent segfault error 4
Message-ID: <041aa0fb00af03026d71bcb288c84db7.NginxMailingListEnglish@forum.nginx.org>

Hi,

I've noticed some nginx segfaults over the last week or so with 1.0.1
and 1.0.2. I havent tested any other version for long enough.
The segfaults are not immediate and sometimes it requires several hours
to reproduce through normal use.

Here is the debug info, if you require anymore information please ask.

[12702.705897] nginx[1769]: segfault at 50545448 ip 0807d1fa sp bfc241d0
error 4 in nginx[8048000+82000]

Linux hostname 2.6.38.6 #1 SMP Sat May 14 09:13:41 UTC 2011 i686
Intel(R) Xeon(TM) CPU 2.40GHz GenuineIntel GNU/Linux

# gcc -v
Reading specs from /usr/lib/gcc/i486-slackware-linux/4.5.2/specs
COLLECT_GCC=gcc
COLLECT_LTO_WRAPPER=/usr/libexec/gcc/i486-slackware-linux/4.5.2/lto-wrapper
Target: i486-slackware-linux
Configured with: ../gcc-4.5.2/configure --prefix=/usr --libdir=/usr/lib
--mandir=/usr/man --infodir=/usr/info --enable-shared --enable-bootstrap
--enable-languages=ada,c,c++,fortran,java,objc,lto
--enable-threads=posix --enable-checking=release --with-system-zlib
--with-python-dir=/lib/python2.6/site-packages
--disable-libunwind-exceptions --enable-__cxa_atexit --enable-libssp
--enable-lto --with-gnu-ld --verbose --with-arch=i486
--target=i486-slackware-linux --build=i486-slackware-linux
--host=i486-slackware-linux
Thread model: posix
gcc version 4.5.2 (GCC)

# ldd /usr/sbin/nginx
        linux-gate.so.1 =>  (0xffffe000)
        libcrypt.so.1 => /lib/libcrypt.so.1 (0xb77ee000)
        libpcre.so.0 => /usr/lib/libpcre.so.0 (0xb77b1000)
        libssl.so.0 => /usr/lib/libssl.so.0 (0xb7769000)
        libcrypto.so.0 => /usr/lib/libcrypto.so.0 (0xb7623000)
        libdl.so.2 => /lib/libdl.so.2 (0xb761e000)
        libz.so.1 => /usr/lib/libz.so.1 (0xb760a000)
        libc.so.6 => /lib/libc.so.6 (0xb74a7000)
        /lib/ld-linux.so.2 (0xb782c000)

glibc-2.13
pcre-8.12
openssl-0.9.8r
zlib-1.2.5

# nginx -V
nginx: nginx version: nginx/1.0.2
nginx: TLS SNI support enabled
nginx: configure arguments: --with-debug --prefix=/usr
--sbin-path=/usr/sbin/nginx --conf-path=/etc/nginx/nginx.conf
--pid-path=/var/run/nginx.pid --lock-path=/var/lock/nginx.lock
--user=www --group=www --error-log-path=/var/log/nginx/error.log
--http-log-path=/var/log/nginx/access.log --with-http_ssl_module
--with-http_realip_module --with-http_addition_module
--with-http_sub_module --with-http_dav_module --with-http_flv_module
--with-http_gzip_static_module --with-http_secure_link_module
--with-http_stub_status_module
--http-client-body-temp-path=/var/lib/nginx/body
--http-proxy-temp-path=/var/lib/nginx/proxy
--http-fastcgi-temp-path=/var/lib/nginx/fastcgi
--http-uwsgi-temp-path=/var/lib/nginx/uwsgi
--http-scgi-temp-path=/var/lib/nginx/scgi --without-mail_pop3_module
--without-mail_imap_module --without-mail_smtp_module

# gdb /usr/sbin/nginx /home/www/core
GNU gdb (GDB) 7.2
Copyright (C) 2010 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later
<http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show
copying"
and "show warranty" for details.
This GDB was configured as "i486-slackware-linux".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/sbin/nginx...(no debugging symbols
found)...done.
[New Thread 1769]

warning: Can't read pathname for load map: Input/output error.
Reading symbols from /lib/libcrypt.so.1...(no debugging symbols
found)...done.
Loaded symbols for /lib/libcrypt.so.1
Reading symbols from /usr/lib/libpcre.so.0...(no debugging symbols
found)...done.
Loaded symbols for /usr/lib/libpcre.so.0
Reading symbols from /usr/lib/libssl.so.0...(no debugging symbols
found)...done.
Loaded symbols for /usr/lib/libssl.so.0
Reading symbols from /usr/lib/libcrypto.so.0...(no debugging symbols
found)...done.
Loaded symbols for /usr/lib/libcrypto.so.0
Reading symbols from /lib/libdl.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/libdl.so.2
Reading symbols from /usr/lib/libz.so.1...(no debugging symbols
found)...done.
Loaded symbols for /usr/lib/libz.so.1
Reading symbols from /lib/libc.so.6...(no debugging symbols
found)...done.
Loaded symbols for /lib/libc.so.6
Reading symbols from /lib/ld-linux.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/ld-linux.so.2
Reading symbols from /lib/libnss_compat.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/libnss_compat.so.2
Reading symbols from /lib/libnsl.so.1...(no debugging symbols
found)...done.
Loaded symbols for /lib/libnsl.so.1
Reading symbols from /lib/libnss_nis.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/libnss_nis.so.2
Reading symbols from /lib/libnss_files.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/libnss_files.so.2
Core was generated by `nginx: worker process                   '.
Program terminated with signal 11, Segmentation fault.
#0  0x0807d1fa in ?? ()
(gdb) backtrace
#0  0x0807d1fa in ?? ()
#1  0x080854ac in ?? ()
#2  0x0807d0c2 in ?? ()
#3  0x08062f13 in ?? ()
#4  0x08062a39 in ?? ()
#5  0x0806a205 in ?? ()
#6  0x080687c1 in ?? ()
#7  0x0806a7cc in ?? ()
#8  0x0806aee7 in ?? ()
#9  0x0804cea8 in ?? ()
#10 0xb7491db6 in __libc_start_main () from /lib/libc.so.6
#11 0x0804bc21 in ?? ()
(gdb) bt full
#0  0x0807d1fa in ?? ()
No symbol table info available.
#1  0x080854ac in ?? ()
No symbol table info available.
#2  0x0807d0c2 in ?? ()
No symbol table info available.
#3  0x08062f13 in ?? ()
No symbol table info available.
#4  0x08062a39 in ?? ()
No symbol table info available.
#5  0x0806a205 in ?? ()
No symbol table info available.
#6  0x080687c1 in ?? ()
No symbol table info available.
#7  0x0806a7cc in ?? ()
No symbol table info available.
#8  0x0806aee7 in ?? ()
No symbol table info available.
#9  0x0804cea8 in ?? ()
No symbol table info available.
#10 0xb7491db6 in __libc_start_main () from /lib/libc.so.6
No symbol table info available.
#11 0x0804bc21 in ?? ()
No symbol table info available.
(gdb) quit


# stat core
  File: `core'
  Size: 1515520         Blocks: 2488       IO Block: 4096   regular
file
Device: 807h/2055d      Inode: 18350096    Links: 1
Access: (0600/-rw-------)  Uid: (   48/     www)   Gid: (   48/    
www)
Access: 2011-05-14 14:08:39.000000000 +0000
Modify: 2011-05-14 14:08:39.000000000 +0000
Change: 2011-05-14 14:08:39.000000000 +0000
 Birth: -

/var/log/nginx# cat segfault.log|grep "14:08:39"
2011/05/14 14:08:39 [debug] 1768#0: timer delta: 501
2011/05/14 14:08:39 [debug] 1768#0: posted events 00000000
2011/05/14 14:08:39 [debug] 1768#0: worker cycle
2011/05/14 14:08:39 [debug] 1768#0: accept mutex lock failed: 0
2011/05/14 14:08:39 [debug] 1768#0: epoll timer: 500
2011/05/14 14:08:39 [debug] 1769#0: epoll: fd:7 ev:0001 d:08D42248
2011/05/14 14:08:39 [debug] 1769#0: *28081 post event 08D5AF50
2011/05/14 14:08:39 [debug] 1769#0: timer delta: 1797
2011/05/14 14:08:39 [debug] 1769#0: posted events 08D5AF50
2011/05/14 14:08:39 [debug] 1769#0: posted event 08D5AF50
2011/05/14 14:08:39 [debug] 1769#0: *28081 delete posted event 08D5AF50
2011/05/14 14:08:39 [debug] 1769#0: *28081 malloc: 08D31F70:680
2011/05/14 14:08:39 [debug] 1769#0: *28081 malloc: 08D33B18:1024
2011/05/14 14:08:39 [debug] 1769#0: *28081 posix_memalign: 08CFAE70:4096
@16
2011/05/14 14:08:39 [debug] 1769#0: *28081 http process request line
2011/05/14 14:08:39 [debug] 1769#0: *28081 recv: fd:7 536 of 1024
2011/05/14 14:08:39 [debug] 1769#0: *28081 http request line: "POST
/index.php?taigachat/list.json HTTP/1.1"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http uri: "/index.php"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http args:
"taigachat/list.json"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http exten: "php"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http process request header
line
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header: "Host:
example.com"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header: "Connection:
keep-alive"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header: "Referer:
http://example.com/forum/"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header: "Content-Length:
142"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header: "Origin:
http://example.com"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header:
"X-Requested-With: XMLHttpRequest"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header: "User-Agent:
Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/534.24 (KHTML, like
Gecko) Chrome/11.0.696.65 Safari/534.24"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header: "Content-Type:
application/x-www-form-urlencoded"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header: "Accept:
application/json, text/javascript, */*; q=0.01"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header:
"Accept-Encoding: gzip,deflate"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header:
"Accept-Language: da-DK,da;q=0.8,en-US;q=0.6,en;q=0.4"
2011/05/14 14:08:39 [debug] 1769#0: *28081 recv: fd:7 -1 of 488
2011/05/14 14:08:39 [debug] 1769#0: *28081 recv() not ready (11:
Resource temporarily unavailable)
2011/05/14 14:08:39 [debug] 1769#0: posted event 00000000
2011/05/14 14:08:39 [debug] 1769#0: worker cycle
2011/05/14 14:08:39 [debug] 1769#0: accept mutex locked
2011/05/14 14:08:39 [debug] 1769#0: epoll timer: 58203
2011/05/14 14:08:39 [debug] 1769#0: epoll: fd:7 ev:0001 d:08D42248
2011/05/14 14:08:39 [debug] 1769#0: *28081 post event 08D5AF50
2011/05/14 14:08:39 [debug] 1769#0: timer delta: 128
2011/05/14 14:08:39 [debug] 1769#0: posted events 08D5AF50
2011/05/14 14:08:39 [debug] 1769#0: posted event 08D5AF50
2011/05/14 14:08:39 [debug] 1769#0: *28081 delete posted event 08D5AF50
2011/05/14 14:08:39 [debug] 1769#0: *28081 http process request header
line
2011/05/14 14:08:39 [debug] 1769#0: *28081 recv: fd:7 488 of 488
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header: "Accept-Charset:
ISO-8859-1,utf-8;q=0.7,*;q=0.3"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header: "Cookie:
__utmz=206500500.1304337103.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none);
xf_user=1931%2C8b0bad949f0d38b81a46469cb9a941ec05485539;
xf_session=53cda77e55ba8e987ba0fb8183cdd326;
__utma=206500500.1451942054.1304337103.1305342568.1305355474.64;
__utmc=206500500; __utmb=206500500.4.10.1305355474"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http header done
2011/05/14 14:08:39 [debug] 1769#0: *28081 event timer del: 7:
4007086959
2011/05/14 14:08:39 [debug] 1769#0: *28081 generic phase: 0
2011/05/14 14:08:39 [debug] 1769#0: *28081 rewrite phase: 1
2011/05/14 14:08:39 [debug] 1769#0: *28081 test location: "/"
2011/05/14 14:08:39 [debug] 1769#0: *28081 test location: "forum/"
2011/05/14 14:08:39 [debug] 1769#0: *28081 test location: ~
"\.(jpg|jpeg|gif|css|png|js|ico)$"
2011/05/14 14:08:39 [debug] 1769#0: *28081 test location: ~ "\.php$"
2011/05/14 14:08:39 [debug] 1769#0: *28081 using configuration "\.php$"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http cl:142 max:52428800
2011/05/14 14:08:39 [debug] 1769#0: *28081 rewrite phase: 3
2011/05/14 14:08:39 [debug] 1769#0: *28081 post rewrite phase: 4
2011/05/14 14:08:39 [debug] 1769#0: *28081 generic phase: 5
2011/05/14 14:08:39 [debug] 1769#0: *28081 generic phase: 6
2011/05/14 14:08:39 [debug] 1769#0: *28081 generic phase: 7
2011/05/14 14:08:39 [debug] 1769#0: *28081 access phase: 8
2011/05/14 14:08:39 [debug] 1769#0: *28081 access phase: 9
2011/05/14 14:08:39 [debug] 1769#0: *28081 post access phase: 10
2011/05/14 14:08:39 [debug] 1769#0: *28081 try files phase: 11
2011/05/14 14:08:39 [debug] 1769#0: *28081 http client request body
preread 137
2011/05/14 14:08:39 [debug] 1769#0: *28081 http read client request
body
2011/05/14 14:08:39 [debug] 1769#0: *28081 recv: fd:7 5 of 5
2011/05/14 14:08:39 [debug] 1769#0: *28081 http client request body recv
5
2011/05/14 14:08:39 [debug] 1769#0: *28081 http client request body rest
0
2011/05/14 14:08:39 [debug] 1769#0: *28081 http init upstream, client
timer: 0
2011/05/14 14:08:39 [debug] 1769#0: *28081 epoll add event: fd:7 op:3
ev:80000005
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"SCRIPT_FILENAME"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var:
"/home/www/example.com/htdocs"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var:
"/index.php"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"SCRIPT_FILENAME: /home/www/example.com/htdocs/index.php"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"QUERY_STRING"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var:
"taigachat/list.json"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "QUERY_STRING:
taigachat/list.json"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"REQUEST_METHOD"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var: "POST"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"REQUEST_METHOD: POST"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"CONTENT_TYPE"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var:
"application/x-www-form-urlencoded"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "CONTENT_TYPE:
application/x-www-form-urlencoded"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"CONTENT_LENGTH"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var: "142"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"CONTENT_LENGTH: 142"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"SCRIPT_NAME"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var:
"/index.php"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "SCRIPT_NAME:
/index.php"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"REQUEST_URI"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var:
"/index.php?taigachat/list.json"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "REQUEST_URI:
/index.php?taigachat/list.json"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"DOCUMENT_URI"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var:
"/index.php"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "DOCUMENT_URI:
/index.php"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"DOCUMENT_ROOT"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var:
"/home/www/example.com/htdocs"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"DOCUMENT_ROOT: /home/www/example.com/htdocs"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"SERVER_PROTOCOL"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var: "HTTP/1.1"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"SERVER_PROTOCOL: HTTP/1.1"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"GATEWAY_INTERFACE"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy: "CGI/1.1"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"GATEWAY_INTERFACE: CGI/1.1"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"SERVER_SOFTWARE"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy: "nginx/"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var: "1.0.2"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"SERVER_SOFTWARE: nginx/1.0.2"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"REMOTE_ADDR"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var: "[REMOTE IP
ADDRESS]"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "REMOTE_ADDR:
[REMOTE IP ADDRESS]"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"REMOTE_PORT"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var: "58057"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "REMOTE_PORT:
58057"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"SERVER_ADDR"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var: "[SERVER IP
ADDRESS]"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "SERVER_ADDR:
[SERVER IP ADDRESS]"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"SERVER_PORT"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var: "80"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "SERVER_PORT:
80"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"SERVER_NAME"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script var:
"example.com"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "SERVER_NAME:
example.com"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy:
"REDIRECT_STATUS"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http script copy: "200"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"REDIRECT_STATUS: 200"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "HTTP_HOST:
example.com"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"HTTP_CONNECTION: keep-alive"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "HTTP_REFERER:
http://example.com/forum/"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"HTTP_CONTENT_LENGTH: 142"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "HTTP_ORIGIN:
http://example.com"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"HTTP_X_REQUESTED_WITH: XMLHttpRequest"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"HTTP_USER_AGENT: Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/534.24
(KHTML, like Gecko) Chrome/11.0.696.65 Safari/534.24"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"HTTP_CONTENT_TYPE: application/x-www-form-urlencoded"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "HTTP_ACCEPT:
application/json, text/javascript, */*; q=0.01"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"HTTP_ACCEPT_ENCODING: gzip,deflate"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"HTTP_ACCEPT_LANGUAGE: da-DK,da;q=0.8,en-US;q=0.6,en;q=0.4"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param:
"HTTP_ACCEPT_CHARSET: ISO-8859-1,utf-8;q=0.7,*;q=0.3"
2011/05/14 14:08:39 [debug] 1769#0: *28081 fastcgi param: "HTTP_COOKIE:
__utmz=206500500.1304337103.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none);
xf_user=1931%2C8b0bad949f0d38b81a46469cb9a941ec05485539;
xf_session=53cda77e55ba8e987ba0fb8183cdd326;
__utma=206500500.1451942054.1304337103.1305342568.1305355474.64;
__utmc=206500500; __utmb=206500500.4.10.1305355474"
2011/05/14 14:08:39 [debug] 1769#0: *28081 posix_memalign: 08CFD490:4096
@16
2011/05/14 14:08:39 [debug] 1769#0: *28081 http cleanup add: 08CFD510
2011/05/14 14:08:39 [debug] 1769#0: *28081 get rr peer, try: 1
2011/05/14 14:08:39 [debug] 1769#0: *28081 socket 9
2011/05/14 14:08:39 [debug] 1769#0: *28081 epoll add connection: fd:9
ev:80000005
2011/05/14 14:08:39 [debug] 1769#0: *28081 connect to
unix:/var/run/php-fpm.sock, fd:9 #28082
2011/05/14 14:08:39 [debug] 1769#0: *28081 connected
2011/05/14 14:08:39 [debug] 1769#0: *28081 http upstream connect: 0
2011/05/14 14:08:39 [debug] 1769#0: *28081 http upstream send request
2011/05/14 14:08:39 [debug] 1769#0: *28081 chain writer buf fl:0 s:1448
2011/05/14 14:08:39 [debug] 1769#0: *28081 chain writer buf fl:0 s:137
2011/05/14 14:08:39 [debug] 1769#0: *28081 chain writer buf fl:0 s:15
2011/05/14 14:08:39 [debug] 1769#0: *28081 chain writer buf fl:0 s:5
2011/05/14 14:08:39 [debug] 1769#0: *28081 chain writer buf fl:0 s:11
2011/05/14 14:08:39 [debug] 1769#0: *28081 chain writer in: 08CFD554
2011/05/14 14:08:39 [debug] 1769#0: *28081 writev: 1616
2011/05/14 14:08:39 [debug] 1769#0: *28081 chain writer out: 00000000
2011/05/14 14:08:39 [debug] 1769#0: *28081 event timer add: 9:
180000:4007208884
2011/05/14 14:08:39 [debug] 1769#0: *28081 http finalize request: -4,
"/index.php?taigachat/list.json" a:1, c:2
2011/05/14 14:08:39 [debug] 1769#0: *28081 http request count:2 blk:0
2011/05/14 14:08:39 [debug] 1769#0: posted event 00000000
2011/05/14 14:08:39 [debug] 1769#0: worker cycle
2011/05/14 14:08:39 [debug] 1769#0: accept mutex locked
2011/05/14 14:08:39 [debug] 1769#0: epoll timer: 180000
2011/05/14 14:08:39 [debug] 1769#0: epoll: fd:7 ev:0005 d:08D42248
2011/05/14 14:08:39 [debug] 1769#0: *28081 post event 08D5AF50
2011/05/14 14:08:39 [debug] 1769#0: *28081 post event 08D67F58
2011/05/14 14:08:39 [debug] 1769#0: epoll: fd:9 ev:0004 d:08D4243D
2011/05/14 14:08:39 [debug] 1769#0: *28081 post event 08D6805C
2011/05/14 14:08:39 [debug] 1769#0: timer delta: 1
2011/05/14 14:08:39 [debug] 1769#0: posted events 08D6805C
2011/05/14 14:08:39 [debug] 1769#0: posted event 08D6805C
2011/05/14 14:08:39 [debug] 1769#0: *28081 delete posted event 08D6805C
2011/05/14 14:08:39 [debug] 1769#0: *28081 http upstream request:
"/index.php?taigachat/list.json"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http upstream dummy handler
2011/05/14 14:08:39 [debug] 1769#0: posted event 08D67F58
2011/05/14 14:08:39 [debug] 1769#0: *28081 delete posted event 08D67F58
2011/05/14 14:08:39 [debug] 1769#0: *28081 http run request:
"/index.php?taigachat/list.json"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http request empty handler
2011/05/14 14:08:39 [debug] 1769#0: posted event 08D5AF50
2011/05/14 14:08:39 [debug] 1769#0: *28081 delete posted event 08D5AF50
2011/05/14 14:08:39 [debug] 1769#0: *28081 http run request:
"/index.php?taigachat/list.json"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http read client request
body
2011/05/14 14:08:39 [debug] 1769#0: *28081 add cleanup: 08CFD63C
2011/05/14 14:08:39 [debug] 1769#0: *28081 hashed path:
/var/lib/nginx/body/0000000001
2011/05/14 14:08:39 [debug] 1769#0: *28081 temp fd:13
2011/05/14 14:08:39 [warn] 1769#0: *28081 a client request body is
buffered to a temporary file /var/lib/nginx/body/0000000001 while
sending request to upstream, client: [REMOTE IP ADDRESS], server:
example.com, request: "POST /index.php?taigachat/list.json HTTP/1.1",
upstream: "fastcgi://unix:/var/run/php-fpm.sock:", host: "example.com",
referrer: "http://example.com/forum/"
2011/05/14 14:08:39 [debug] 1769#0: *28081 write: 13, 08CFB790, 5, 0
2011/05/14 14:08:39 [debug] 1769#0: *28081 recv: fd:7 0 of 0
2011/05/14 14:08:39 [debug] 1769#0: *28081 http client request body recv
0
2011/05/14 14:08:39 [info] 1769#0: *28081 client closed prematurely
connection while sending request to upstream, client: [REMOTE IP
ADDRESS], server: example.com, request: "POST
/index.php?taigachat/list.json HTTP/1.1", upstream:
"fastcgi://unix:/var/run/php-fpm.sock:", host: "example.com", referrer:
"http://example.com/forum/"
2011/05/14 14:08:39 [debug] 1769#0: *28081 http finalize request: 400,
"/index.php?taigachat/list.json" a:1, c:1
2011/05/14 14:08:39 [debug] 1769#0: *28081 http terminate request
count:1
2011/05/14 14:08:39 [debug] 1769#0: *28081 cleanup http upstream
request: "/index.php"
2011/05/14 14:08:39 [debug] 1769#0: *28081 finalize http upstream
request: -4
2011/05/14 14:08:39 [debug] 1769#0: *28081 finalize http fastcgi
request
2011/05/14 14:08:39 [debug] 1769#0: *28081 free rr peer 1 0
2011/05/14 14:08:39 [debug] 1769#0: *28081 close http upstream
connection: 9
2011/05/14 14:08:39 [debug] 1769#0: *28081 event timer del: 9:
4007208884
2011/05/14 14:08:39 [debug] 1769#0: *28081 reusable connection: 0
2011/05/14 14:08:39 [debug] 1769#0: *28081 http finalize request: -4,
"/index.php?taigachat/list.json" a:1, c:1
2011/05/14 14:08:39 [debug] 1769#0: *28081 set http keepalive handler
2011/05/14 14:08:39 [debug] 1769#0: *28081 http close request
2011/05/14 14:08:39 [debug] 1769#0: *28081 http log handler
2011/05/14 14:08:39 [debug] 1769#0: *28081 run cleanup: 08CFD63C
2011/05/14 14:08:39 [debug] 1769#0: *28081 file cleanup: fd:13
2011/05/14 14:08:39 [debug] 1769#0: *28081 free: 08CFAE70, unused: 5
2011/05/14 14:08:39 [debug] 1769#0: *28081 free: 08CFD490, unused: 3417
2011/05/14 14:08:39 [debug] 1769#0: *28081 event timer add: 7:
75000:4007103885
2011/05/14 14:08:39 [debug] 1769#0: *28081 free: 08D31F70
2011/05/14 14:08:39 [debug] 1769#0: *28081 free: 08D33B18
2011/05/14 14:08:39 [debug] 1769#0: *28081 hc free: 00000000 0
2011/05/14 14:08:39 [debug] 1769#0: *28081 hc busy: 00000000 0
2011/05/14 14:08:39 [debug] 1769#0: *28081 tcp_nodelay
2011/05/14 14:08:39 [debug] 1769#0: *28081 reusable connection: 1
2011/05/14 14:08:39 [notice] 1767#0: signal 17 (SIGCHLD) received
2011/05/14 14:08:39 [alert] 1767#0: worker process 1769 exited on signal
11 (core dumped)
2011/05/14 14:08:39 [debug] 1767#0: wake up, sigio 0
2011/05/14 14:08:39 [debug] 1767#0: reap children
2011/05/14 14:08:39 [debug] 1767#0: child: 0 1768 e:0 t:0 d:0 r:1 j:0
2011/05/14 14:08:39 [debug] 1767#0: child: 1 1769 e:0 t:1 d:0 r:1 j:0
2011/05/14 14:08:39 [debug] 1767#0: pass close channel s:1 pid:1769
to:1768
2011/05/14 14:08:39 [debug] 1767#0: channel 7:10
2011/05/14 14:08:39 [debug] 1768#0: epoll: fd:9 ev:0001 d:08D41C6C
2011/05/14 14:08:39 [debug] 1768#0: channel handler
2011/05/14 14:08:39 [debug] 1768#0: channel: 16
2011/05/14 14:08:39 [debug] 1768#0: channel command: 2
2011/05/14 14:08:39 [debug] 1768#0: close channel s:1 pid:1769 our:1769
fd:3
2011/05/14 14:08:39 [debug] 1768#0: channel: -2
2011/05/14 14:08:39 [debug] 1768#0: timer delta: 376
2011/05/14 14:08:39 [debug] 1768#0: posted events 00000000
2011/05/14 14:08:39 [debug] 1768#0: worker cycle
2011/05/14 14:08:39 [debug] 1768#0: accept mutex locked
2011/05/14 14:08:39 [debug] 1768#0: epoll add event: fd:8 op:1
ev:00000001
2011/05/14 14:08:39 [debug] 1768#0: epoll timer: 21259
2011/05/14 14:08:39 [notice] 1767#0: start worker process 3483
2011/05/14 14:08:39 [debug] 1767#0: pass channel s:1 pid:3483 fd:7 to
s:0 pid:1768 fd:3
2011/05/14 14:08:39 [debug] 1767#0: sigsuspend
2011/05/14 14:08:39 [debug] 1768#0: epoll: fd:9 ev:0001 d:08D41C6C
2011/05/14 14:08:39 [debug] 1768#0: post event 08D5AC44
2011/05/14 14:08:39 [debug] 1768#0: timer delta: 0
2011/05/14 14:08:39 [debug] 1768#0: posted events 08D5AC44
2011/05/14 14:08:39 [debug] 1768#0: posted event 08D5AC44
2011/05/14 14:08:39 [notice] 1767#0: signal 29 (SIGIO) received
2011/05/14 14:08:39 [debug] 1768#0: delete posted event 08D5AC44
2011/05/14 14:08:39 [debug] 1767#0: wake up, sigio 0
2011/05/14 14:08:39 [debug] 1768#0: channel handler
2011/05/14 14:08:39 [debug] 1767#0: sigsuspend
2011/05/14 14:08:39 [debug] 1768#0: channel: 16
2011/05/14 14:08:39 [debug] 1768#0: channel command: 1
2011/05/14 14:08:39 [debug] 1768#0: get channel s:1 pid:3483 fd:3
2011/05/14 14:08:39 [debug] 1768#0: channel: -2
2011/05/14 14:08:39 [debug] 1768#0: posted event 00000000
2011/05/14 14:08:39 [debug] 1768#0: worker cycle
2011/05/14 14:08:39 [debug] 1768#0: accept mutex locked
2011/05/14 14:08:39 [debug] 1768#0: epoll timer: 21259
2011/05/14 14:08:39 [notice] 3483#0: sched_setaffinity(0x0000000A)
2011/05/14 14:08:39 [debug] 3483#0: malloc: 08D0A9F8:6144
2011/05/14 14:08:39 [debug] 3483#0: malloc: 08D41C08:102400
2011/05/14 14:08:39 [debug] 3483#0: malloc: 08D5AC10:53248
2011/05/14 14:08:39 [debug] 3483#0: malloc: 08D67C18:53248
2011/05/14 14:08:39 [debug] 3483#0: epoll add event: fd:10 op:1
ev:00000001
2011/05/14 14:08:39 [debug] 3483#0: setproctitle: "nginx: worker
process"
2011/05/14 14:08:39 [debug] 3483#0: worker cycle
2011/05/14 14:08:39 [debug] 3483#0: accept mutex lock failed: 0
2011/05/14 14:08:39 [debug] 3483#0: epoll timer: 500


# cat nginx.conf
user  www www;
worker_processes  2;
worker_cpu_affinity 0101 1010;
error_log  /var/log/nginx/segfault.log  debug;
#error_log  /var/log/nginx/error.log  notice;

worker_rlimit_core  500M;
working_directory   /home/www/;

events {
    worker_connections  1024;
}
http {
    include       mime.types;
    default_type  application/octet-stream;
    log_format  upstream '$remote_addr - - [$time_local] "$request"
$status '
                         'upstream $upstream_response_time request
$request_time bytes $bytes_sent body_bytes $body_bytes_sent '
                         '[$host]';

    sendfile on;
    server_tokens off;
    keepalive_timeout 75;
    client_max_body_size 50M;

    gzip on;
    gzip_proxied any;
    gzip_vary on;
    gzip_disable "msie6";
    gzip_types text/plain text/css application/json
application/x-javascript text/xml application/xml application/xml+rss
text/javascript;

    fastcgi_max_temp_file_size 0;
    open_file_cache max=1000 inactive=120s;
    open_file_cache_valid    120s;
    open_file_cache_min_uses 5;
    open_file_cache_errors   on;

    index  index.php index.html;
    include /etc/nginx/enabled/*.conf;
}

# cat enabled/default.conf
server {
   listen 80 default;
   server_name  _;
   server_name_in_redirect off;

   log_format  vhost  '$host $remote_addr - $remote_user [$time_local]
"$request" '
                      '$status $body_bytes_sent "$http_referer" '
                      '"$http_user_agent" "$http_x_forwarded_for"';

   access_log  /var/log/nginx/default_vhost.log vhost;
   root   /home/www/default/;

   location ~ \.php$ {
       try_files      $uri =404;
       fastcgi_pass   unix:/var/run/php-fpm.sock;
       fastcgi_index  index.php;
       include        fastcgi.conf;
       fastcgi_intercept_errors on;
   }

   location ~* \.(jpg|jpeg|gif|css|png|js|ico)$ {
       log_not_found off;
       access_log    off;
       expires       max;
   }
}

# cat enabled/example.com.conf
server {
   server_name www.example.com;
   rewrite ^ $scheme://example.com$request_uri? permanent;
}

server {
   server_name  example.com;
   root         /home/www/example.com/htdocs/;

   location ~* \.(jpg|jpeg|gif|css|png|js|ico)$ {
       access_log off;
       expires    30d;
   }

   location ^~ /forum/ {
       rewrite  ^/forum/$      /$1?  last;
       rewrite  ^/forum/(.+)$  /$1?  permanent;
   }

   location / {
       try_files $uri $uri/ /index.php?$uri&$args;
   }

   location /blog/ {
       try_files $uri $uri/ /blog/index.php?$uri&$args;
   }

   location ~ \.php$ {
        try_files      $uri =404;
        access_log     /var/log/nginx/upstream.log  upstream;
        fastcgi_pass   unix:/var/run/php-fpm.sock;
        fastcgi_index  index.php;
        include        fastcgi.conf;
        fastcgi_intercept_errors        on;
        fastcgi_ignore_client_abort     on;
   }

   location ~ /\.ht {
         access_log    off;
         log_not_found off;
         deny all;
   }

   location ~ /(internal_data|library)/ {
       allow   127.0.0.1;
       deny    all;
   }
}

# cat localhost.conf
server {
   server_name  localhost 127.0.0.1;
   access_log  off;
   root   /home/www/default/;
   location /nginx_status {      
      stub_status on;
      access_log  off;
      allow       127.0.0.1;
      deny        all;
   }
}

# cat fastcgi.conf

fastcgi_param  SCRIPT_FILENAME    $document_root$fastcgi_script_name;
fastcgi_param  QUERY_STRING       $query_string;
fastcgi_param  REQUEST_METHOD     $request_method;
fastcgi_param  CONTENT_TYPE       $content_type;
fastcgi_param  CONTENT_LENGTH     $content_length;

fastcgi_param  SCRIPT_NAME        $fastcgi_script_name;
fastcgi_param  REQUEST_URI        $request_uri;
fastcgi_param  DOCUMENT_URI       $document_uri;
fastcgi_param  DOCUMENT_ROOT      $document_root;
fastcgi_param  SERVER_PROTOCOL    $server_protocol;

fastcgi_param  GATEWAY_INTERFACE  CGI/1.1;
fastcgi_param  SERVER_SOFTWARE    nginx/$nginx_version;

fastcgi_param  REMOTE_ADDR        $remote_addr;
fastcgi_param  REMOTE_PORT        $remote_port;
fastcgi_param  SERVER_ADDR        $server_addr;
fastcgi_param  SERVER_PORT        $server_port;
fastcgi_param  SERVER_NAME        $server_name;

# PHP only, required if PHP was built with --enable-force-cgi-redirect
fastcgi_param  REDIRECT_STATUS    200;






Also an earlier segfault, however I was not running a error_log at debug
at the time. Same nginx build and same config as above.

# nginx -V
nginx: nginx version: nginx/1.0.2
nginx: TLS SNI support enabled
nginx: configure arguments: --with-debug --prefix=/usr
--sbin-path=/usr/sbin/nginx --conf-path=/etc/nginx/nginx.conf
--pid-path=/var/run/nginx.pid --lock-path=/var/lock/nginx.lock
--user=www --group=www --error-log-path=/var/log/nginx/error.log
--http-log-path=/var/log/nginx/access.log --with-http_ssl_module
--with-http_realip_module --with-http_addition_module
--with-http_sub_module --with-http_dav_module --with-http_flv_module
--with-http_gzip_static_module --with-http_secure_link_module
--with-http_stub_status_module
--http-client-body-temp-path=/var/lib/nginx/body
--http-proxy-temp-path=/var/lib/nginx/proxy
--http-fastcgi-temp-path=/var/lib/nginx/fastcgi
--http-uwsgi-temp-path=/var/lib/nginx/uwsgi
--http-scgi-temp-path=/var/lib/nginx/scgi --without-mail_pop3_module
--without-mail_imap_module --without-mail_smtp_module

# stat core
  File: `core'
  Size: 1490944         Blocks: 2232       IO Block: 4096   regular
file
Device: 807h/2055d      Inode: 18350095    Links: 1
Access: (0600/-rw-------)  Uid: (   48/     www)   Gid: (   48/    
www)
Access: 2011-05-14 10:14:35.000000000 +0000
Modify: 2011-05-14 10:14:35.000000000 +0000
Change: 2011-05-14 10:14:35.000000000 +0000
 Birth: -

# gdb /usr/sbin/nginx /home/www/core
GNU gdb (GDB) 7.2
Copyright (C) 2010 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later
<http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show
copying"
and "show warranty" for details.
This GDB was configured as "i486-slackware-linux".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/sbin/nginx...(no debugging symbols
found)...done.
[New Thread 18354]

warning: Can't read pathname for load map: Input/output error.
Reading symbols from /lib/libcrypt.so.1...(no debugging symbols
found)...done.
Loaded symbols for /lib/libcrypt.so.1
Reading symbols from /usr/lib/libpcre.so.0...(no debugging symbols
found)...done.
Loaded symbols for /usr/lib/libpcre.so.0
Reading symbols from /usr/lib/libssl.so.0...(no debugging symbols
found)...done.
Loaded symbols for /usr/lib/libssl.so.0
Reading symbols from /usr/lib/libcrypto.so.0...(no debugging symbols
found)...done.
Loaded symbols for /usr/lib/libcrypto.so.0
Reading symbols from /lib/libdl.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/libdl.so.2
Reading symbols from /usr/lib/libz.so.1...(no debugging symbols
found)...done.
Loaded symbols for /usr/lib/libz.so.1
Reading symbols from /lib/libc.so.6...(no debugging symbols
found)...done.
Loaded symbols for /lib/libc.so.6
Reading symbols from /lib/ld-linux.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/ld-linux.so.2
Reading symbols from /lib/libnss_compat.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/libnss_compat.so.2
Reading symbols from /lib/libnsl.so.1...(no debugging symbols
found)...done.
Loaded symbols for /lib/libnsl.so.1
Reading symbols from /lib/libnss_nis.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/libnss_nis.so.2
Reading symbols from /lib/libnss_files.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/libnss_files.so.2
Core was generated by `nginx:'.
Program terminated with signal 11, Segmentation fault.
#0  0x0807d1fa in ?? ()
(gdb) bt
#0  0x0807d1fa in ?? ()
#1  0x080854ac in ?? ()
#2  0x0807d0c2 in ?? ()
#3  0x08062f13 in ?? ()
#4  0x08062a39 in ?? ()
#5  0x0806a205 in ?? ()
#6  0x080687c1 in ?? ()
#7  0x0806a7cc in ?? ()
#8  0x0806aee7 in ?? ()
#9  0x0804cea8 in ?? ()
#10 0xb74afdb6 in __libc_start_main () from /lib/libc.so.6
#11 0x0804bc21 in ?? ()
(gdb) bt full
#0  0x0807d1fa in ?? ()
No symbol table info available.
#1  0x080854ac in ?? ()
No symbol table info available.
#2  0x0807d0c2 in ?? ()
No symbol table info available.
#3  0x08062f13 in ?? ()
No symbol table info available.
#4  0x08062a39 in ?? ()
No symbol table info available.
#5  0x0806a205 in ?? ()
No symbol table info available.
#6  0x080687c1 in ?? ()
No symbol table info available.
#7  0x0806a7cc in ?? ()
No symbol table info available.
#8  0x0806aee7 in ?? ()
No symbol table info available.
#9  0x0804cea8 in ?? ()
No symbol table info available.
#10 0xb74afdb6 in __libc_start_main () from /lib/libc.so.6
No symbol table info available.
#11 0x0804bc21 in ?? ()
No symbol table info available.
(gdb) quit

# error.log
2011/05/14 10:14:35 [warn] 18354#0: *1707 a client request body is
buffered to a temporary file /var/lib/nginx/body/0000000001 while
sending request to upstream, client: [CLIENT IP ADDRESS], server:
example.com, request: "POST /login/csrf-token-refresh HTTP/1.1",
upstream: "fastcgi://unix:/var/run/php-fpm.sock:", host: "example.com",
referrer: "http://example.com/threads/THREADNAME.9709/"
2011/05/14 10:14:35 [info] 18354#0: *1707 client closed prematurely
connection while sending request to upstream, client: [CLIENT IP
ADDRESS], server: example.com, request: "POST /login/csrf-token-refresh
HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm.sock:", host:
"example.com", referrer: "http://example.com/threads/THREADNAME.9709/"
2011/05/14 10:14:35 [notice] 18353#0: signal 17 (SIGCHLD) received
2011/05/14 10:14:35 [alert] 18353#0: worker process 18354 exited on
signal 11 (core dumped)
2011/05/14 10:14:35 [notice] 18353#0: start worker process 14632
2011/05/14 10:14:35 [notice] 18353#0: signal 29 (SIGIO) received
2011/05/14 10:14:35 [notice] 14632#0: sched_setaffinity(0x0000000A)

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198003,198003#msg-198003



From ionathan at gmail.com  Sat May 14 19:28:37 2011
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Sat, 14 May 2011 12:28:37 -0300
Subject: nginx worker processes
Message-ID: <BANLkTimC78iWs8HjrnfYnKEE3Dbw4fXfcQ@mail.gmail.com>

Is it true to say that every worker process in nginx has its own round-robin
counters and that they are not aware of the other workers counters?

Thanks!

Jonathan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110514/f870332f/attachment.html>

From nginx-forum at nginx.us  Sat May 14 19:46:12 2011
From: nginx-forum at nginx.us (jjjx128)
Date: Sat, 14 May 2011 11:46:12 -0400
Subject: nginx worker processes
In-Reply-To: <BANLkTimC78iWs8HjrnfYnKEE3Dbw4fXfcQ@mail.gmail.com>
References: <BANLkTimC78iWs8HjrnfYnKEE3Dbw4fXfcQ@mail.gmail.com>
Message-ID: <8e337b6dcb17affd78b39a148b4fa1bf.NginxMailingListEnglish@forum.nginx.org>

Jonathan Leibiusky Wrote:
-------------------------------------------------------
> Is it true to say that every worker process in
> nginx has its own round-robin
> counters and that they are not aware of the other
> workers counters?

Yes, take a look at ngx_http_upstream_round_robin.c.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198006,198008#msg-198008



From mp3geek at gmail.com  Sat May 14 19:48:19 2011
From: mp3geek at gmail.com (Ryan B)
Date: Sun, 15 May 2011 03:48:19 +1200
Subject: seperating ipv6 and ipv4 traffic
Message-ID: <BANLkTi=+50UA9kCmLTTkNMBt-MguhK-NjQ@mail.gmail.com>

server {
        server_name secure.mydomain.com;
        listen       443;
        gzip         on;
        gzip_comp_level  1;
        gzip_types      text/plain text/html text/css
application/x-javascript text/xml text/javascript;
        gzip_static on;
        gzip_http_version   1.1;
        gzip_proxied        any;
        gzip_disable        "msie6";
        gzip_vary           on;
        ssl          on;
        ssl_ciphers RC4:ALL:-LOW:-EXPORT:!ADH:!MD5;
        keepalive_timeout  0;
        ssl_certificate      /root/server.pem;
        ssl_certificate_key  /root/ssl.key;

           location / {
                root   /var/www;
                index  index.html index.htm index.php;
            }

        }

Seems to work well (and accepts both SSL ipv4/ipv6 since they point to
same ip)..
But I'd like to move ipv4-only traffic away to another ip. How is this
done? I'm guessing I create 2 of these, specifying ipv4 in one and
ipv6 in the other?


From mp3geek at gmail.com  Sat May 14 19:58:01 2011
From: mp3geek at gmail.com (Ryan B)
Date: Sun, 15 May 2011 03:58:01 +1200
Subject: seperating ipv6 and ipv4 traffic
In-Reply-To: <BANLkTi=+50UA9kCmLTTkNMBt-MguhK-NjQ@mail.gmail.com>
References: <BANLkTi=+50UA9kCmLTTkNMBt-MguhK-NjQ@mail.gmail.com>
Message-ID: <BANLkTim=M2ayK=8dQ7VUm18=HUYCtxGiQg@mail.gmail.com>

Ah, also, with using redirect with a request SSL then redirected to a
non-ssl site work?
On May 15, 2011 3:48 AM, "Ryan B" <mp3geek at gmail.com> wrote:
> server {
> server_name secure.mydomain.com;
> listen 443;
> gzip on;
> gzip_comp_level 1;
> gzip_types text/plain text/html text/css
> application/x-javascript text/xml text/javascript;
> gzip_static on;
> gzip_http_version 1.1;
> gzip_proxied any;
> gzip_disable "msie6";
> gzip_vary on;
> ssl on;
> ssl_ciphers RC4:ALL:-LOW:-EXPORT:!ADH:!MD5;
> keepalive_timeout 0;
> ssl_certificate /root/server.pem;
> ssl_certificate_key /root/ssl.key;
>
> location / {
> root /var/www;
> index index.html index.htm index.php;
> }
>
> }
>
> Seems to work well (and accepts both SSL ipv4/ipv6 since they point to
> same ip)..
> But I'd like to move ipv4-only traffic away to another ip. How is this
> done? I'm guessing I create 2 of these, specifying ipv4 in one and
> ipv6 in the other?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110515/5d5394ee/attachment.html>

From nginx-forum at nginx.us  Sat May 14 20:22:08 2011
From: nginx-forum at nginx.us (jfix)
Date: Sat, 14 May 2011 12:22:08 -0400
Subject: gzip compression not taken into account?
In-Reply-To: <a1bb7953bf979c5dba6b2802f7e5b67a.NginxMailingListEnglish@forum.nginx.org>
References: <a1bb7953bf979c5dba6b2802f7e5b67a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <0c89e375c17616d0cdc09c2d633194dd.NginxMailingListEnglish@forum.nginx.org>

Thanks for your replies, I will try to provide more information (I'm
quite new with nginx).

Traffic passes through the company firewall/proxy server, so potentially
things could get mangled there.  I guess  there is no way to inspect the
requested headers as received by the server? I set the error_log level
to debug_http, but it doesn't seem to log headers.

One interesting detail: Using curl and explicitly requesting a
compressed response like this, I get this result:

M:>curl -v --compressed http://example.org/
* About to connect() to example.org port 80 (#0)
*   Trying xx.xx.xx.xx... connected
* Connected to example.org (xx.xx.xx.xx) port 80 (#0)
> GET / HTTP/1.1
> User-Agent: curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.8l
zlib/1.2.3
> Host: example.org
> Accept: */*
> Accept-Encoding: deflate, gzip
>
< HTTP/1.1 200 OK
< Server: nginx
< Date: Sat, 14 May 2011 16:10:06 GMT
< Content-Type: text/html; charset=UTF-8
< Transfer-Encoding: chunked
< Connection: keep-alive
< Vary: Accept-Encoding
< Content-Encoding: gzip
<
10.107.145.6 - - [14/May/2011:18:10:06 +0000] "GET / HTTP/1.1" 200 2671
"-" "curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.81
zlib/1.2.3 "-" "0.031" "2.53"

where the last item "2.53" is the compression ratio.

M:>curl -v --compressed http://example.org/resource/style/global.css
* About to connect() to example.org port 80 (#0)
*   Trying xx.xx.xx.xx... connected
* Connected to example.org (xx.xx.xx.xx) port 80 (#0)
> GET /resource/style/global.css HTTP/1.1
> User-Agent: curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.8l
zlib/1.2.3
> Host: example.org
> Accept: */*
> Accept-Encoding: deflate, gzip
>
< HTTP/1.1 200 OK
< Server: nginx
< Date: Sat, 14 May 2011 16:14:38 GMT
< Content-Type: text/css
< Transfer-Encoding: chunked
< Connection: keep-alive
< Vary: Accept-Encoding
< Last-Modified: Fri, 13 May 2011 16:00:36 GMT
< Expires: Thu, 31 Dec 2037 23:55:55 GMT
< Cache-Control: max-age=315360000
< Content-Encoding: gzip

10.107.145.6 - - [14/May/2011:18:14:38 +0000] "GET / HTTP/1.1" 200 7917
"-" "curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.81
zlib/1.2.3 "-" "0.000" "5.48"

So, this means that compression does work in theory, but it doesn't
unless I request it. The following example is the same as the previous
one, but without the --compressed:

M:>curl -v http://example.org/resource/style/global.css
* About to connect() to example.org port 80 (#0)
*   Trying xx.xx.xx.xx... connected
* Connected to example.org (xx.xx.xx.xx) port 80 (#0)
> GET /resource/style/global.css HTTP/1.1
> User-Agent: curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.8l
zlib/1.2.3
> Host: example.org
> Accept: */*
>
< HTTP/1.1 200 OK
< Server: nginx
< Date: Sat, 14 May 2011 16:16:59 GMT
< Content-Type: text/css
< Connection: keep-alive
< Vary: Accept-Encoding
< Content-Length: 43345
< Last-Modified: Fri, 13 May 2011 16:00:36 GMT
< Expires: Thu, 31 Dec 2037 23:55:55 GMT
< Cache-Control: max-age=315360000

10.107.145.6 - - [14/May/2011:18:16:59 +0000] "GET / HTTP/1.1" 200 43345
"-" "curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.81
zlib/1.2.3 "-" "0.000" "-"

I'm not sure what the --compressed option translates to in terms of
headers, but it makes a difference. The last examples is what I get when
I user a browser (Chrome and Firefox tested).

Thanks for your help,
Jakob.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197898,198024#msg-198024



From mat999 at gmail.com  Sat May 14 20:23:30 2011
From: mat999 at gmail.com (SplitIce)
Date: Sun, 15 May 2011 02:23:30 +1000
Subject: seperating ipv6 and ipv4 traffic
In-Reply-To: <BANLkTim=M2ayK=8dQ7VUm18=HUYCtxGiQg@mail.gmail.com>
References: <BANLkTi=+50UA9kCmLTTkNMBt-MguhK-NjQ@mail.gmail.com>
	<BANLkTim=M2ayK=8dQ7VUm18=HUYCtxGiQg@mail.gmail.com>
Message-ID: <BANLkTi=OZLhbU5oEHcF7BeqbuXVg-L4FuQ@mail.gmail.com>

listen on the IPv6 addr.

On Sun, May 15, 2011 at 1:58 AM, Ryan B <mp3geek at gmail.com> wrote:

> Ah, also, with using redirect with a request SSL then redirected to a
> non-ssl site work?
> On May 15, 2011 3:48 AM, "Ryan B" <mp3geek at gmail.com> wrote:
> > server {
> > server_name secure.mydomain.com;
> > listen 443;
> > gzip on;
> > gzip_comp_level 1;
> > gzip_types text/plain text/html text/css
> > application/x-javascript text/xml text/javascript;
> > gzip_static on;
> > gzip_http_version 1.1;
> > gzip_proxied any;
> > gzip_disable "msie6";
> > gzip_vary on;
> > ssl on;
> > ssl_ciphers RC4:ALL:-LOW:-EXPORT:!ADH:!MD5;
> > keepalive_timeout 0;
> > ssl_certificate /root/server.pem;
> > ssl_certificate_key /root/ssl.key;
> >
> > location / {
> > root /var/www;
> > index index.html index.htm index.php;
> > }
> >
> > }
> >
> > Seems to work well (and accepts both SSL ipv4/ipv6 since they point to
> > same ip)..
> > But I'd like to move ipv4-only traffic away to another ip. How is this
> > done? I'm guessing I create 2 of these, specifying ipv4 in one and
> > ipv6 in the other?
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>


-- 
Warez Scene <http://thewarezscene.org> Free Rapidshare
Downloads<http://www.nexusddl.com>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110515/8a754bea/attachment.html>

From ionathan at gmail.com  Sat May 14 21:29:12 2011
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Sat, 14 May 2011 14:29:12 -0300
Subject: nginx worker processes
In-Reply-To: <8e337b6dcb17affd78b39a148b4fa1bf.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTimC78iWs8HjrnfYnKEE3Dbw4fXfcQ@mail.gmail.com>
	<8e337b6dcb17affd78b39a148b4fa1bf.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTim4VNuiH+mXfkH+ybZktCpgahTR1A@mail.gmail.com>

Thanks, it's been a while since I last developed in c, so I wanted
someone else's opinion.
I saw that there is a fair round-robin module and I also saw that it
uses shared memory to store something. Does it mean that using that
module counters and servers state (either it is up or down) are
shared?

On 5/14/11, jjjx128 <nginx-forum at nginx.us> wrote:
> Jonathan Leibiusky Wrote:
> -------------------------------------------------------
>> Is it true to say that every worker process in
>> nginx has its own round-robin
>> counters and that they are not aware of the other
>> workers counters?
>
> Yes, take a look at ngx_http_upstream_round_robin.c.
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,198006,198008#msg-198008
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>


From tomlove at gmail.com  Sun May 15 01:36:10 2011
From: tomlove at gmail.com (Thomas Love)
Date: Sat, 14 May 2011 23:36:10 +0200
Subject: gzip compression not taken into account?
In-Reply-To: <0c89e375c17616d0cdc09c2d633194dd.NginxMailingListEnglish@forum.nginx.org>
References: <a1bb7953bf979c5dba6b2802f7e5b67a.NginxMailingListEnglish@forum.nginx.org>
	<0c89e375c17616d0cdc09c2d633194dd.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTinfkRvKV=na0Ce=0e-qGzz2Bx5yjA@mail.gmail.com>

The --compressed flag is adding "Accept-Encoding: deflate, gzip" to the
request.

Your test shows that nginx is working fine. It is compressing the content if
it receives an Accept-Encoding: gzip field, and doesn't if not. This is
valid behaviour.

However, Chrome *should* be sending an Accept-Encoding field just like curl
--compressed. You can confirm this using the dev tools. If it is, and you
don't get a Content-Encoding: gzip header in the response on Chrome
(indicating compression), and if that curl test you performed was run on
local interface to nginx, then it is almost certainly an HTTP proxy that's
interfering.

If otoh, you got those results running the curl test remotely so that it too
went through the proxy (a useful confirmation test if you didn't), then I'm
not so sure.


Cheers
Thomas


On 14 May 2011 18:22, jfix <nginx-forum at nginx.us> wrote:

> Thanks for your replies, I will try to provide more information (I'm
> quite new with nginx).
>
> Traffic passes through the company firewall/proxy server, so potentially
> things could get mangled there.  I guess  there is no way to inspect the
> requested headers as received by the server? I set the error_log level
> to debug_http, but it doesn't seem to log headers.
>
> One interesting detail: Using curl and explicitly requesting a
> compressed response like this, I get this result:
>
> M:>curl -v --compressed http://example.org/
> * About to connect() to example.org port 80 (#0)
> *   Trying xx.xx.xx.xx... connected
> * Connected to example.org (xx.xx.xx.xx) port 80 (#0)
> > GET / HTTP/1.1
> > User-Agent: curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.8l
> zlib/1.2.3
> > Host: example.org
> > Accept: */*
> > Accept-Encoding: deflate, gzip
> >
> < HTTP/1.1 200 OK
> < Server: nginx
> < Date: Sat, 14 May 2011 16:10:06 GMT
> < Content-Type: text/html; charset=UTF-8
> < Transfer-Encoding: chunked
> < Connection: keep-alive
> < Vary: Accept-Encoding
> < Content-Encoding: gzip
> <
> 10.107.145.6 - - [14/May/2011:18:10:06 +0000] "GET / HTTP/1.1" 200 2671
> "-" "curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.81
> zlib/1.2.3 "-" "0.031" "2.53"
>
> where the last item "2.53" is the compression ratio.
>
> M:>curl -v --compressed http://example.org/resource/style/global.css
> * About to connect() to example.org port 80 (#0)
> *   Trying xx.xx.xx.xx... connected
> * Connected to example.org (xx.xx.xx.xx) port 80 (#0)
> > GET /resource/style/global.css HTTP/1.1
> > User-Agent: curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.8l
> zlib/1.2.3
> > Host: example.org
> > Accept: */*
> > Accept-Encoding: deflate, gzip
> >
> < HTTP/1.1 200 OK
> < Server: nginx
> < Date: Sat, 14 May 2011 16:14:38 GMT
> < Content-Type: text/css
> < Transfer-Encoding: chunked
> < Connection: keep-alive
> < Vary: Accept-Encoding
> < Last-Modified: Fri, 13 May 2011 16:00:36 GMT
> < Expires: Thu, 31 Dec 2037 23:55:55 GMT
> < Cache-Control: max-age=315360000
> < Content-Encoding: gzip
>
> 10.107.145.6 - - [14/May/2011:18:14:38 +0000] "GET / HTTP/1.1" 200 7917
> "-" "curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.81
> zlib/1.2.3 "-" "0.000" "5.48"
>
> So, this means that compression does work in theory, but it doesn't
> unless I request it. The following example is the same as the previous
> one, but without the --compressed:
>
> M:>curl -v http://example.org/resource/style/global.css
> * About to connect() to example.org port 80 (#0)
> *   Trying xx.xx.xx.xx... connected
> * Connected to example.org (xx.xx.xx.xx) port 80 (#0)
> > GET /resource/style/global.css HTTP/1.1
> > User-Agent: curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.8l
> zlib/1.2.3
> > Host: example.org
> > Accept: */*
> >
> < HTTP/1.1 200 OK
> < Server: nginx
> < Date: Sat, 14 May 2011 16:16:59 GMT
> < Content-Type: text/css
> < Connection: keep-alive
> < Vary: Accept-Encoding
> < Content-Length: 43345
> < Last-Modified: Fri, 13 May 2011 16:00:36 GMT
> < Expires: Thu, 31 Dec 2037 23:55:55 GMT
> < Cache-Control: max-age=315360000
>
> 10.107.145.6 - - [14/May/2011:18:16:59 +0000] "GET / HTTP/1.1" 200 43345
> "-" "curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.81
> zlib/1.2.3 "-" "0.000" "-"
>
> I'm not sure what the --compressed option translates to in terms of
> headers, but it makes a difference. The last examples is what I get when
> I user a browser (Chrome and Firefox tested).
>
> Thanks for your help,
> Jakob.
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,197898,198024#msg-198024
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110514/def367b7/attachment-0001.html>

From nginx-forum at nginx.us  Sun May 15 01:38:49 2011
From: nginx-forum at nginx.us (the_traveller)
Date: Sat, 14 May 2011 17:38:49 -0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <2507a2f6fd2c5e8a743f33ca869dfa82.NginxMailingListEnglish@forum.nginx.org>
References: <72a81ff3e80ed0b4b78cb81a348826cf@ruby-forum.com>
	<2507a2f6fd2c5e8a743f33ca869dfa82.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <bde4d0b65344b7647dfb70a9edb30ea5.NginxMailingListEnglish@forum.nginx.org>

;(

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,23868,198081#msg-198081



From nginx-forum at nginx.us  Sun May 15 01:51:33 2011
From: nginx-forum at nginx.us (jfix)
Date: Sat, 14 May 2011 17:51:33 -0400
Subject: gzip compression not taken into account?
In-Reply-To: <0c89e375c17616d0cdc09c2d633194dd.NginxMailingListEnglish@forum.nginx.org>
References: <a1bb7953bf979c5dba6b2802f7e5b67a.NginxMailingListEnglish@forum.nginx.org>
	<0c89e375c17616d0cdc09c2d633194dd.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <e1eb6064bdd055b2c4aca6d4471692ca.NginxMailingListEnglish@forum.nginx.org>

Thanks Thomas for your reply.

I've redone the test by explicitly supplying the proxy, and indeed it
seems to remove the Accept-encoding header, as even with --compressed
option, the response is now uncompressed:

M:>curl --compressed -v -x wsg-proxy.example.org:80
http://example.org/resource/style/global.css
* About to connect() to proxy wsg-proxy.example.org port 80 (#0)
*   Trying 10.101.128.15... connected
* Connected to wsg-proxy.oecd.org (10.101.128.15) port 80 (#0)
> GET http://example.org/resource/style/global.css HTTP/1.1
> User-Agent: curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.8l
zlib/1.2.3
> Host: example.org
> Accept: */*
> Accept-Encoding: deflate, gzip
> Proxy-Connection: Keep-Alive
>
< HTTP/1.1 200 OK
< Server: nginx
< Date: Sat, 14 May 2011 21:46:49 GMT
< Content-Type: text/css
< Vary: Accept-Encoding
< Last-Modified: Fri, 13 May 2011 16:00:36 GMT
< Expires: Thu, 31 Dec 2037 23:55:55 GMT
< Cache-Control: max-age=315360000
< Content-Length: 43345
< Age: 0
< Via: 1.1 localhost.localdomain

So, it looks like I have to talk to my company's IT people ... Thanks
for your help.

Jakob.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197898,198083#msg-198083



From tomlove at gmail.com  Sun May 15 03:25:40 2011
From: tomlove at gmail.com (Thomas Love)
Date: Sun, 15 May 2011 01:25:40 +0200
Subject: nginx 1.0.2 intermittent segfault error 4
In-Reply-To: <041aa0fb00af03026d71bcb288c84db7.NginxMailingListEnglish@forum.nginx.org>
References: <041aa0fb00af03026d71bcb288c84db7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTiktGzhhAriZHDfjSPiZXinFPYpFEw@mail.gmail.com>

Hmmm - I'm curious: what's the value of the backtrace without symbol table?
How to get one for nginx?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110515/81b54f4d/attachment.html>

From nginx-forum at nginx.us  Sun May 15 07:46:41 2011
From: nginx-forum at nginx.us (holstebm)
Date: Sat, 14 May 2011 23:46:41 -0400
Subject: add_header works in server but not in http?
Message-ID: <bf5bfe80b68f51516b864d624c5ccf89.NginxMailingListEnglish@forum.nginx.org>

Hello,  I decided I'd clean up my configuration files for Nginx.  It's
long overdue.

I have the following in my nginx.conf but is never sent.  I don't have
another reference to this header anywhere.

add_header                                X-UA-Compatible
'IE=edge,chrome=1';

Moreover, if I move the line into a Server block, it works fine.

Docs say can be used in Http, Server and Location.  I'm at a loss.

Bret

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198143,198143#msg-198143



From nginx-forum at nginx.us  Sun May 15 07:58:26 2011
From: nginx-forum at nginx.us (holstebm)
Date: Sat, 14 May 2011 23:58:26 -0400
Subject: add_header works in server but not in http?
In-Reply-To: <bf5bfe80b68f51516b864d624c5ccf89.NginxMailingListEnglish@forum.nginx.org>
References: <bf5bfe80b68f51516b864d624c5ccf89.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <2e52965b4dcd75e917b186825edf2bac.NginxMailingListEnglish@forum.nginx.org>

Okay, then, a little more detail.  In my Server block, I had an
add_header for Cache-control.  When I remove that add_header, the other
one in Http appears.  If I put both in Http they both appear.  If I put
both in Server they both appear but it seems that the add_header removes
the headers from the parent blocks.  I don't think that was the previous
behavior.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198143,198147#msg-198147



From rschultz7 at gmail.com  Sun May 15 15:20:06 2011
From: rschultz7 at gmail.com (Rob Schultz)
Date: Sun, 15 May 2011 13:20:06 +0200
Subject: IPv6 issues viewing any static or dynamic file
Message-ID: <20024305-77F2-45ED-9D1D-D189407A1C0E@gmail.com>

Hello,
	I am having issues with IPv6 being able to view any files static or dynamic. I keep getting the error 
curl -6 http://ipv6.example.com/test.png

2011/05/14 16:18:36 [info] 6217#0: *1 client timed out (110: Connection timed out) while sending response to client, client: xxxx:xxx:xxxx:xxxx:xxx:xxxx:xxxx:xxxx, server: example.com, request: "GET /test.png HTTP/1.1", host: "example.com"

but i can do HEAD requests and such and get a response. 

 curl -6 -I ipv6.example.com
HTTP/1.1 200 OK
Server: nginx/1.0.2
Date: Sun, 15 May 2011 08:44:31 GMT
Content-Type: text/html
Content-Length: 0
Last-Modified: Thu, 24 Jan 2008 19:27:58 GMT
Connection: keep-alive
Keep-Alive: timeout=10
Accept-Ranges: bytes

nginx -V
nginx: nginx version: nginx/1.0.2
nginx: TLS SNI support enabled
nginx: configure arguments: --prefix=/usr --sbin-path=/usr/sbin/nginx --conf-path=/etc/nginx/nginx.conf --error-log-path=/var/log/nginx/error_log --pid-path=/var/run/nginx.pid --lock-path=/var/lock/nginx.lock --user=nginx --group=nginx --with-cc-opt=-I/usr/include --with-ld-opt=-L/usr/lib --http-log-path=/var/log/nginx/access_log --http-client-body-temp-path=/var/tmp/nginx/client --http-proxy-temp-path=/var/tmp/nginx/proxy --http-fastcgi-temp-path=/var/tmp/nginx/fastcgi --http-scgi-temp-path=/var/tmp/nginx/scgi --http-uwsgi-temp-path=/var/tmp/nginx/uwsgi --with-ipv6 --with-pcre --with-http_gzip_static_module --with-http_realip_module --with-http_ssl_module --without-mail_imap_module --without-mail_pop3_module --without-mail_smtp_module

and nginx.conf

user nginx nginx;
worker_processes 2;
worker_rlimit_nofile 8192;

error_log /var/log/nginx/error_log info;
#error_log /var/log/nginx/error_log debug;
events {
        worker_connections  8192;
#       worker_connections  1024;
        use epoll;
}

http {
        include         /etc/nginx/mime.types;
#       include         /etc/nginx/proxy.conf;
        include         /etc/nginx/fastcgi_params;

        default_type    application/octet-stream;

        log_format main
                '$remote_addr - $remote_user [$time_local] '
                '"$request" $status $bytes_sent '
                '"$http_referer" "$http_user_agent" '
                '"$gzip_ratio"';

        client_header_timeout   10;
        client_body_timeout     10;
        #needed for file uploads > 1m
        client_max_body_size    10m;
        send_timeout            10;

        connection_pool_size            256;
        client_header_buffer_size       1k;
        large_client_header_buffers     4 2k;
        request_pool_size               4k;

        gzip on;
        gzip_min_length 1100;
        gzip_buffers    16 8k;
        gzip_static     on;
        gzip_types      text/plain text/css application/x-javascript text/xml application/xml application/xml+rss text/javascript;

        output_buffers  1 32k;
        postpone_output 1460;

        sendfile        on;
        tcp_nopush      on;
        tcp_nodelay     on;

        keepalive_timeout       10 10;

        ignore_invalid_headers  on;

        index index.html index.php;
        server {
                listen          [::]:80 default ipv6only=on;
                listen          80 default;
                server_name     example.com www.example.com ipv6.example.com;

                access_log      /var/log/nginx/texample.com.access_log main;
                error_log       /var/log/nginx/example.com.error_log info;
        
                root /var/www/example.com/htdocs;
                autoindex on;

                location /blog/ {
                        if (!-e $request_filename) {
                                rewrite  ^.+?(/.*\.php)$  /blog$1          last;
                                rewrite  ^.+?(/wp-.*)     /blog$1          last;
                                rewrite  ^                /blog/index.php  last;
                        }
                }
        
                location ~ ^.+\.php {
                        include /etc/nginx/fastcgi_params;
                        fastcgi_param SCRIPT_FILENAME /var/www/example.com/htdocs$fastcgi_script_name;
                        fastcgi_pass   127.0.0.1:9000;
                        fastcgi_index   index.php;
                }
        
                location ~* ^.+\.(jpg|jpeg|gif|png)$ {
                        access_log   off;
                        expires      30d;
                }

                location ^~ /c/ {
                        return 403;
                }
                location ^~ /p/ {
                        return 403;
                }
        }
        include /etc/nginx/vhosts.d/*.conf;
}

i have turned firewall off to ensure that wasn't an issue. I am not sure were else to look. 

Here is also a debug log
http://pastie.org/private/dbzwwlkxaelmex47hvwnzq

v/r,
Rob
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110515/0832a278/attachment.html>

From igor at sysoev.ru  Sun May 15 16:26:27 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Sun, 15 May 2011 16:26:27 +0400
Subject: IPv6 issues viewing any static or dynamic file
In-Reply-To: <20024305-77F2-45ED-9D1D-D189407A1C0E@gmail.com>
References: <20024305-77F2-45ED-9D1D-D189407A1C0E@gmail.com>
Message-ID: <20110515122627.GA79835@sysoev.ru>

On Sun, May 15, 2011 at 01:20:06PM +0200, Rob Schultz wrote:
> Hello,
> 	I am having issues with IPv6 being able to view any files static or dynamic. I keep getting the error 
> curl -6 http://ipv6.example.com/test.png
> 
> 2011/05/14 16:18:36 [info] 6217#0: *1 client timed out (110: Connection timed out) while sending response to client, client: xxxx:xxx:xxxx:xxxx:xxx:xxxx:xxxx:xxxx, server: example.com, request: "GET /test.png HTTP/1.1", host: "example.com"
> 
> but i can do HEAD requests and such and get a response. 
> 
>  curl -6 -I ipv6.example.com
> HTTP/1.1 200 OK
> Server: nginx/1.0.2
> Date: Sun, 15 May 2011 08:44:31 GMT
> Content-Type: text/html
> Content-Length: 0
> Last-Modified: Thu, 24 Jan 2008 19:27:58 GMT
> Connection: keep-alive
> Keep-Alive: timeout=10
> Accept-Ranges: bytes

How do you connect via IPv6 ?
It seems like large packets tunneling issue.
Are you able to scp large file via IPv6 ?


-- 
Igor Sysoev


From igor at sysoev.ru  Sun May 15 16:33:27 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Sun, 15 May 2011 16:33:27 +0400
Subject: seperating ipv6 and ipv4 traffic
In-Reply-To: <BANLkTi=+50UA9kCmLTTkNMBt-MguhK-NjQ@mail.gmail.com>
References: <BANLkTi=+50UA9kCmLTTkNMBt-MguhK-NjQ@mail.gmail.com>
Message-ID: <20110515123327.GB79835@sysoev.ru>

On Sun, May 15, 2011 at 03:48:19AM +1200, Ryan B wrote:
> server {
>         server_name secure.mydomain.com;
>         listen       443;
>         gzip         on;
>         gzip_comp_level  1;
>         gzip_types      text/plain text/html text/css
> application/x-javascript text/xml text/javascript;
>         gzip_static on;
>         gzip_http_version   1.1;
>         gzip_proxied        any;
>         gzip_disable        "msie6";
>         gzip_vary           on;
>         ssl          on;
>         ssl_ciphers RC4:ALL:-LOW:-EXPORT:!ADH:!MD5;
>         keepalive_timeout  0;
>         ssl_certificate      /root/server.pem;
>         ssl_certificate_key  /root/ssl.key;
> 
>            location / {
>                 root   /var/www;
>                 index  index.html index.htm index.php;
>             }
> 
>         }
> 
> Seems to work well (and accepts both SSL ipv4/ipv6 since they point to
> same ip)..
> But I'd like to move ipv4-only traffic away to another ip. How is this
> done? I'm guessing I create 2 of these, specifying ipv4 in one and
> ipv6 in the other?

  server {
      listen  192.168.1.1:443;
      listen  [...]:443;


-- 
Igor Sysoev


From igor at sysoev.ru  Sun May 15 16:34:18 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Sun, 15 May 2011 16:34:18 +0400
Subject: seperating ipv6 and ipv4 traffic
In-Reply-To: <BANLkTim=M2ayK=8dQ7VUm18=HUYCtxGiQg@mail.gmail.com>
References: <BANLkTi=+50UA9kCmLTTkNMBt-MguhK-NjQ@mail.gmail.com>
	<BANLkTim=M2ayK=8dQ7VUm18=HUYCtxGiQg@mail.gmail.com>
Message-ID: <20110515123418.GC79835@sysoev.ru>

On Sun, May 15, 2011 at 03:58:01AM +1200, Ryan B wrote:
> Ah, also, with using redirect with a request SSL then redirected to a
> non-ssl site work?

Sorry, I do not understand the question.


-- 
Igor Sysoev


From igor at sysoev.ru  Sun May 15 16:50:04 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Sun, 15 May 2011 16:50:04 +0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <bde4d0b65344b7647dfb70a9edb30ea5.NginxMailingListEnglish@forum.nginx.org>
References: <72a81ff3e80ed0b4b78cb81a348826cf@ruby-forum.com>
	<2507a2f6fd2c5e8a743f33ca869dfa82.NginxMailingListEnglish@forum.nginx.org>
	<bde4d0b65344b7647dfb70a9edb30ea5.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110515125004.GD79835@sysoev.ru>

On Sat, May 14, 2011 at 05:38:49PM -0400, the_traveller wrote:
> ;(

Rewrite rules are not configuration means. They are primitive programming
language that is too hard to understand. For example, these

RewriteRule ^chat/.*    pages.php [QSA]
RewriteRule ^Songs(.*)$ showcat.php?g=2 [QSA]

may be translated to

location /maindir/chat/ {
    fastcgi_pass   ...
    fastcgi_param  SCRIPT_FILENAME   /path/to/script/pages.php;
    fastcgi_param  QUERY_STRING      $args;
    ...
}

location /maindir/Songs/ {
    fastcgi_pass   ...
    fastcgi_param  SCRIPT_FILENAME   /path/to/script/showcat.php;
    fastcgi_param  QUERY_STRING      g=2$is_args$args;
    ...
}


-- 
Igor Sysoev


From igor at sysoev.ru  Sun May 15 16:52:22 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Sun, 15 May 2011 16:52:22 +0400
Subject: add_header works in server but not in http?
In-Reply-To: <2e52965b4dcd75e917b186825edf2bac.NginxMailingListEnglish@forum.nginx.org>
References: <bf5bfe80b68f51516b864d624c5ccf89.NginxMailingListEnglish@forum.nginx.org>
	<2e52965b4dcd75e917b186825edf2bac.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110515125221.GE79835@sysoev.ru>

On Sat, May 14, 2011 at 11:58:26PM -0400, holstebm wrote:
> Okay, then, a little more detail.  In my Server block, I had an
> add_header for Cache-control.  When I remove that add_header, the other
> one in Http appears.  If I put both in Http they both appear.  If I put
> both in Server they both appear but it seems that the add_header removes
> the headers from the parent blocks.  I don't think that was the previous
> behavior.

add_header is inherited if there is no add_header on the current level.
This behavior was from from the very start.


-- 
Igor Sysoev


From nunomagalhaes at eu.ipp.pt  Sun May 15 17:46:31 2011
From: nunomagalhaes at eu.ipp.pt (=?UTF-8?Q?Nuno_Magalh=C3=A3es?=)
Date: Sun, 15 May 2011 14:46:31 +0100
Subject: Anything wrong with the wiki server?
Message-ID: <BANLkTi=EkZGH6Xgwefc8J8ZZmBeXq8oJZg@mail.gmail.com>

I can connect to nginx.org but not to wiki.nginx.org, times out.


From nginx-forum at nginx.us  Sun May 15 18:42:34 2011
From: nginx-forum at nginx.us (inph)
Date: Sun, 15 May 2011 10:42:34 -0400
Subject: nginx 1.0.2 intermittent segfault error 4
In-Reply-To: <041aa0fb00af03026d71bcb288c84db7.NginxMailingListEnglish@forum.nginx.org>
References: <041aa0fb00af03026d71bcb288c84db7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4c5ea4c84247919300aa51e1226efb52.NginxMailingListEnglish@forum.nginx.org>

Sorry, was in a hurry and didnt even notice.

nginx compiled with -g and -O0 (previously segfaults above were compiled
with -O2)
nginx.conf is the same as before.

Might be worth noting that this segfault took approximately 12 hours to
happen after i recompiled with -g -O0
However this might be down to low traffic between midnight and midday on
Sunday.

# nginx -V
nginx: nginx version: nginx/1.0.2
nginx: TLS SNI support enabled
nginx: configure arguments: --with-debug --prefix=/usr
--sbin-path=/usr/sbin/nginx --conf-path=/etc/nginx/nginx.conf
--pid-path=/var/run/nginx.pid --lock-path=/var/lock/nginx.lock
--user=www --group=www --error-log-path=/var/log/nginx/error.log
--http-log-path=/var/log/nginx/access.log --with-http_ssl_module
--with-http_realip_module --with-http_addition_module
--with-http_sub_module --with-http_dav_module --with-http_flv_module
--with-http_gzip_static_module --with-http_secure_link_module
--with-http_stub_status_module
--http-client-body-temp-path=/var/lib/nginx/body
--http-proxy-temp-path=/var/lib/nginx/proxy
--http-fastcgi-temp-path=/var/lib/nginx/fastcgi
--http-uwsgi-temp-path=/var/lib/nginx/uwsgi
--http-scgi-temp-path=/var/lib/nginx/scgi --without-mail_pop3_module
--without-mail_imap_module --without-mail_smtp_module


# stat core
  File: `core'
  Size: 1359872         Blocks: 2352       IO Block: 4096   regular
file
Device: 807h/2055d      Inode: 18350097    Links: 1
Access: (0600/-rw-------)  Uid: (   48/     www)   Gid: (   48/    
www)
Access: 2011-05-15 12:41:37.000000000 +0000
Modify: 2011-05-15 12:43:32.000000000 +0000
Change: 2011-05-15 12:43:32.000000000 +0000
 Birth: -


# gdb /usr/sbin/nginx /home/www/core
GNU gdb (GDB) 7.2
Copyright (C) 2010 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later
<http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show
copying"
and "show warranty" for details.
This GDB was configured as "i486-slackware-linux".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/sbin/nginx...done.
[New Thread 9301]

warning: Can't read pathname for load map: Input/output error.
Reading symbols from /lib/libcrypt.so.1...(no debugging symbols
found)...done.
Loaded symbols for /lib/libcrypt.so.1
Reading symbols from /usr/lib/libpcre.so.0...(no debugging symbols
found)...done.
Loaded symbols for /usr/lib/libpcre.so.0
Reading symbols from /usr/lib/libssl.so.0...(no debugging symbols
found)...done.
Loaded symbols for /usr/lib/libssl.so.0
Reading symbols from /usr/lib/libcrypto.so.0...(no debugging symbols
found)...done.
Loaded symbols for /usr/lib/libcrypto.so.0
Reading symbols from /lib/libdl.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/libdl.so.2
Reading symbols from /usr/lib/libz.so.1...(no debugging symbols
found)...done.
Loaded symbols for /usr/lib/libz.so.1
Reading symbols from /lib/libc.so.6...(no debugging symbols
found)...done.
Loaded symbols for /lib/libc.so.6
Reading symbols from /lib/ld-linux.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/ld-linux.so.2
Reading symbols from /lib/libnss_compat.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/libnss_compat.so.2
Reading symbols from /lib/libnsl.so.1...(no debugging symbols
found)...done.
Loaded symbols for /lib/libnsl.so.1
Reading symbols from /lib/libnss_nis.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/libnss_nis.so.2
Reading symbols from /lib/libnss_files.so.2...(no debugging symbols
found)...done.
Loaded symbols for /lib/libnss_files.so.2
Core was generated by `nginx:'.
Program terminated with signal 11, Segmentation fault.
#0  0x08085243 in ngx_http_terminate_request (r=0x8985548, rc=400) at
src/http/ngx_http_request.c:2085
2085        ngx_log_debug2(NGX_LOG_DEBUG_HTTP, r->connection->log, 0,
(gdb) backtrace
#0  0x08085243 in ngx_http_terminate_request (r=0x8985548, rc=400) at
src/http/ngx_http_request.c:2085
#1  0x08084c91 in ngx_http_finalize_request (r=0x8985548, rc=400) at
src/http/ngx_http_request.c:1912
#2  0x0808dab7 in ngx_http_read_client_request_body_handler
(r=0x8985548) at src/http/ngx_http_request_body.c:254
#3  0x080849ab in ngx_http_request_handler (ev=0x89ab0ac) at
src/http/ngx_http_request.c:1800
#4  0x08067f7b in ngx_event_process_posted (cycle=0x8946cb8,
posted=0x80eaaf4) at src/event/ngx_event_posted.c:39
#5  0x08066531 in ngx_process_events_and_timers (cycle=0x8946cb8) at
src/event/ngx_event.c:272
#6  0x0807051b in ngx_worker_process_cycle (cycle=0x8946cb8, data=0x0)
at src/os/unix/ngx_process_cycle.c:800
#7  0x0806dc5f in ngx_spawn_process (cycle=0x8946cb8, proc=0x807038a
<ngx_worker_process_cycle>, data=0x0, name=0x80d2e2b "worker process",
respawn=-3)
    at src/os/unix/ngx_process.c:196
#8  0x0806f813 in ngx_start_worker_processes (cycle=0x8946cb8, n=2,
type=-3) at src/os/unix/ngx_process_cycle.c:360
#9  0x0806f05e in ngx_master_process_cycle (cycle=0x8946cb8) at
src/os/unix/ngx_process_cycle.c:136
#10 0x0804c297 in main (argc=1, argv=0xbfdb0fc4) at
src/core/nginx.c:405
(gdb) backtrace full
#0  0x08085243 in ngx_http_terminate_request (r=0x8985548, rc=400) at
src/http/ngx_http_request.c:2085
        cln = 0x0
        mr = 0x8985548
        e = 0x8984dc8
#1  0x08084c91 in ngx_http_finalize_request (r=0x8985548, rc=400) at
src/http/ngx_http_request.c:1912
        c = 0x8992434
        pr = 0x8992434
        clcf = 0x894b804
#2  0x0808dab7 in ngx_http_read_client_request_body_handler
(r=0x8985548) at src/http/ngx_http_request_body.c:254
        rc = 400
#3  0x080849ab in ngx_http_request_handler (ev=0x89ab0ac) at
src/http/ngx_http_request.c:1800
        c = 0x8992434
        r = 0x8985548
        ctx = 0x8984df0
#4  0x08067f7b in ngx_event_process_posted (cycle=0x8946cb8,
posted=0x80eaaf4) at src/event/ngx_event_posted.c:39
        ev = 0x89ab0ac
#5  0x08066531 in ngx_process_events_and_timers (cycle=0x8946cb8) at
src/event/ngx_event.c:272
        flags = 3
        timer = 197
        delta = 0
#6  0x0807051b in ngx_worker_process_cycle (cycle=0x8946cb8, data=0x0)
at src/os/unix/ngx_process_cycle.c:800
        i = 9299
        c = 0xbfdb0cb8
#7  0x0806dc5f in ngx_spawn_process (cycle=0x8946cb8, proc=0x807038a
<ngx_worker_process_cycle>, data=0x0, name=0x80d2e2b "worker process",
respawn=-3)
    at src/os/unix/ngx_process.c:196
        on = 1
        pid = 0
        s = 1
#8  0x0806f813 in ngx_start_worker_processes (cycle=0x8946cb8, n=2,
type=-3) at src/os/unix/ngx_process_cycle.c:360
        i = 1
        ch = {command = 1, pid = 9300, slot = 0, fd = 3}
#9  0x0806f05e in ngx_master_process_cycle (cycle=0x8946cb8) at
src/os/unix/ngx_process_cycle.c:136
        title = 0x898feb6 "master process nginx"
        p = 0x898feca ""
        size = 21
        i = 1
        n = 143944900
        sigio = 0
        set = {__val = {0 <repeats 32 times>}}
        itv = {it_interval = {tv_sec = 135072442, tv_usec = 9299},
it_value = {tv_sec = 0, tv_usec = 0}}
        live = 0
        delay = 0
        ls = 0x0
        ccf = 0x89474b0
#10 0x0804c297 in main (argc=1, argv=0xbfdb0fc4) at
src/core/nginx.c:405
        i = 53
        log = 0x80e88d0
        cycle = 0x8946cb8
        init_cycle = {conf_ctx = 0x0, pool = 0x8945dd0, log = 0x80e88d0,
new_log = {log_level = 0, file = 0x0, connection = 0, handler = 0, data
= 0x0, action = 0x0},
          files = 0x0, free_connections = 0x0, free_connection_n = 0,
reusable_connections_queue = {prev = 0x0, next = 0x0}, listening = {elts
= 0x0, nelts = 0, size = 0,
            nalloc = 0, pool = 0x0}, pathes = {elts = 0x0, nelts = 0,
size = 0, nalloc = 0, pool = 0x0}, open_files = {last = 0x0, part =
{elts = 0x0, nelts = 0, next = 0x0},
            size = 0, nalloc = 0, pool = 0x0}, shared_memory = {last =
0x0, part = {elts = 0x0, nelts = 0, next = 0x0}, size = 0, nalloc = 0,
pool = 0x0}, connection_n = 0,
          files_n = 0, connections = 0x0, read_events = 0x0,
write_events = 0x0, old_cycle = 0x0, conf_file = {len = 21, data =
0x80cf5c2 "/etc/nginx/nginx.conf"},
          conf_param = {len = 0, data = 0x0}, conf_prefix = {len = 11,
data = 0x80cf5c2 "/etc/nginx/nginx.conf"}, prefix = {len = 5, data =
0x80cf5bc "/usr/"}, lock_file = {
            len = 0, data = 0x0}, hostname = {len = 0, data = 0x0}}
        ccf = 0x89474b0
(gdb) quit


[93995.665429] nginx[9301]: segfault at 50545448 ip 08085243 sp bfdb0b50
error 4 in nginx[8048000+97000]

# cat segfault.log |grep "12:43:32"
2011/05/15 12:43:32 [debug] 9812#0: timer delta: 501
2011/05/15 12:43:32 [debug] 9812#0: posted events 00000000
2011/05/15 12:43:32 [debug] 9812#0: worker cycle
2011/05/15 12:43:32 [debug] 9812#0: accept mutex lock failed: 0
2011/05/15 12:43:32 [debug] 9812#0: epoll timer: 500
2011/05/15 12:43:32 [debug] 9301#0: epoll: fd:10 ev:0001 d:08992435
2011/05/15 12:43:32 [debug] 9301#0: *4558 post event 089AB0AC
2011/05/15 12:43:32 [debug] 9301#0: timer delta: 801
2011/05/15 12:43:32 [debug] 9301#0: posted events 089AB0AC
2011/05/15 12:43:32 [debug] 9301#0: posted event 089AB0AC
2011/05/15 12:43:32 [debug] 9301#0: *4558 delete posted event 089AB0AC
2011/05/15 12:43:32 [debug] 9301#0: *4558 malloc: 08985548:680
2011/05/15 12:43:32 [debug] 9301#0: *4558 malloc: 08983960:1024
2011/05/15 12:43:32 [debug] 9301#0: *4558 posix_memalign: 0894AF70:4096
@16
2011/05/15 12:43:32 [debug] 9301#0: *4558 http process request line
2011/05/15 12:43:32 [debug] 9301#0: *4558 recv: fd:10 536 of 1024
2011/05/15 12:43:32 [debug] 9301#0: *4558 http request line: "POST
/index.php?taigachat/list.json HTTP/1.1"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http uri: "/index.php"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http args:
"taigachat/list.json"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http exten: "php"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http process request header
line
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header: "Host:
example.com"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header: "Connection:
keep-alive"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header: "Referer:
http://example.com/forum/"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header: "Content-Length:
142"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header: "Origin:
http://example.com"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header:
"X-Requested-With: XMLHttpRequest"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header: "User-Agent:
Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/534.24 (KHTML, like
Gecko) Chrome/11.0.696.65 Safari/534.24"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header: "Content-Type:
application/x-www-form-urlencoded"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header: "Accept:
application/json, text/javascript, */*; q=0.01"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header: "Accept-Encoding:
gzip,deflate"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header: "Accept-Language:
da-DK,da;q=0.8,en-US;q=0.6,en;q=0.4"
2011/05/15 12:43:32 [debug] 9301#0: *4558 recv: fd:10 -1 of 488
2011/05/15 12:43:32 [debug] 9301#0: *4558 recv() not ready (11: Resource
temporarily unavailable)
2011/05/15 12:43:32 [debug] 9301#0: posted event 00000000
2011/05/15 12:43:32 [debug] 9301#0: worker cycle
2011/05/15 12:43:32 [debug] 9301#0: accept mutex locked
2011/05/15 12:43:32 [debug] 9301#0: epoll timer: 210
2011/05/15 12:43:32 [debug] 9301#0: epoll: fd:10 ev:0001 d:08992435
2011/05/15 12:43:32 [debug] 9301#0: *4558 post event 089AB0AC
2011/05/15 12:43:32 [debug] 9301#0: timer delta: 12
2011/05/15 12:43:32 [debug] 9301#0: posted events 089AB0AC
2011/05/15 12:43:32 [debug] 9301#0: posted event 089AB0AC
2011/05/15 12:43:32 [debug] 9301#0: *4558 delete posted event 089AB0AC
2011/05/15 12:43:32 [debug] 9301#0: *4558 http process request header
line
2011/05/15 12:43:32 [debug] 9301#0: *4558 recv: fd:10 488 of 488
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header: "Accept-Charset:
ISO-8859-1,utf-8;q=0.7,*;q=0.3"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header: "Cookie:
__utmz=206500500.1304337103.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none);
xf_user=1931%2C8b0bad949f0d38b81a46469cb9a941ec05485539;
xf_session=094d80a983afbbd15ac06937d4b23798;
__utma=206500500.1451942054.1304337103.1305433729.1305437755.69;
__utmc=206500500; __utmb=206500500.3.10.1305437755"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http header done
2011/05/15 12:43:32 [debug] 9301#0: *4558 event timer del: 10:
4088381035
2011/05/15 12:43:32 [debug] 9301#0: *4558 generic phase: 0
2011/05/15 12:43:32 [debug] 9301#0: *4558 rewrite phase: 1
2011/05/15 12:43:32 [debug] 9301#0: *4558 test location: "/"
2011/05/15 12:43:32 [debug] 9301#0: *4558 test location: "forum/"
2011/05/15 12:43:32 [debug] 9301#0: *4558 test location: ~
"\.(jpg|jpeg|gif|css|png|js|ico)$"
2011/05/15 12:43:32 [debug] 9301#0: *4558 test location: ~ "\.php$"
2011/05/15 12:43:32 [debug] 9301#0: *4558 using configuration "\.php$"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http cl:142 max:52428800
2011/05/15 12:43:32 [debug] 9301#0: *4558 rewrite phase: 3
2011/05/15 12:43:32 [debug] 9301#0: *4558 post rewrite phase: 4
2011/05/15 12:43:32 [debug] 9301#0: *4558 generic phase: 5
2011/05/15 12:43:32 [debug] 9301#0: *4558 generic phase: 6
2011/05/15 12:43:32 [debug] 9301#0: *4558 generic phase: 7
2011/05/15 12:43:32 [debug] 9301#0: *4558 access phase: 8
2011/05/15 12:43:32 [debug] 9301#0: *4558 access phase: 9
2011/05/15 12:43:32 [debug] 9301#0: *4558 post access phase: 10
2011/05/15 12:43:32 [debug] 9301#0: *4558 try files phase: 11
2011/05/15 12:43:32 [debug] 9301#0: *4558 http client request body
preread 137
2011/05/15 12:43:32 [debug] 9301#0: *4558 http read client request body
2011/05/15 12:43:32 [debug] 9301#0: *4558 recv: fd:10 5 of 5
2011/05/15 12:43:32 [debug] 9301#0: *4558 http client request body recv
5
2011/05/15 12:43:32 [debug] 9301#0: *4558 http client request body rest
0
2011/05/15 12:43:32 [debug] 9301#0: *4558 http init upstream, client
timer: 0
2011/05/15 12:43:32 [debug] 9301#0: *4558 epoll add event: fd:10 op:3
ev:80000005
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"SCRIPT_FILENAME"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var:
"/home/www/example.com/htdocs"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var: "/index.php"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"SCRIPT_FILENAME: /home/www/example.com/htdocs/index.php"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"QUERY_STRING"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var:
"taigachat/list.json"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "QUERY_STRING:
taigachat/list.json"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"REQUEST_METHOD"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var: "POST"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"REQUEST_METHOD: POST"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"CONTENT_TYPE"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var:
"application/x-www-form-urlencoded"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "CONTENT_TYPE:
application/x-www-form-urlencoded"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"CONTENT_LENGTH"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var: "142"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"CONTENT_LENGTH: 142"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"SCRIPT_NAME"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var: "/index.php"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "SCRIPT_NAME:
/index.php"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"REQUEST_URI"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var:
"/index.php?taigachat/list.json"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "REQUEST_URI:
/index.php?taigachat/list.json"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"DOCUMENT_URI"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var: "/index.php"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "DOCUMENT_URI:
/index.php"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"DOCUMENT_ROOT"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var:
"/home/www/example.com/htdocs"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "DOCUMENT_ROOT:
/home/www/example.com/htdocs"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"SERVER_PROTOCOL"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var: "HTTP/1.1"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"SERVER_PROTOCOL: HTTP/1.1"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"GATEWAY_INTERFACE"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy: "CGI/1.1"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"GATEWAY_INTERFACE: CGI/1.1"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"SERVER_SOFTWARE"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy: "nginx/"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var: "1.0.2"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"SERVER_SOFTWARE: nginx/1.0.2"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"REMOTE_ADDR"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var: "[CLIENT IP
ADDRESS]"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "REMOTE_ADDR:
[CLIENT IP ADDRESS]"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"REMOTE_PORT"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var: "50294"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "REMOTE_PORT:
50294"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"SERVER_ADDR"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var: "[SERVER IP
ADDRESS]"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "SERVER_ADDR:
[SERVER IP ADDRESS]"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"SERVER_PORT"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var: "80"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "SERVER_PORT:
80"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"SERVER_NAME"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script var:
"example.com"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "SERVER_NAME:
example.com"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy:
"REDIRECT_STATUS"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http script copy: "200"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"REDIRECT_STATUS: 200"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "HTTP_HOST:
example.com"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"HTTP_CONNECTION: keep-alive"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "HTTP_REFERER:
http://example.com/forum/"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"HTTP_CONTENT_LENGTH: 142"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "HTTP_ORIGIN:
http://example.com"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"HTTP_X_REQUESTED_WITH: XMLHttpRequest"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"HTTP_USER_AGENT: Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/534.24
(KHTML, like Gecko) Chrome/11.0.696.65 Safari/534.24"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"HTTP_CONTENT_TYPE: application/x-www-form-urlencoded"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "HTTP_ACCEPT:
application/json, text/javascript, */*; q=0.01"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"HTTP_ACCEPT_ENCODING: gzip,deflate"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"HTTP_ACCEPT_LANGUAGE: da-DK,da;q=0.8,en-US;q=0.6,en;q=0.4"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param:
"HTTP_ACCEPT_CHARSET: ISO-8859-1,utf-8;q=0.7,*;q=0.3"
2011/05/15 12:43:32 [debug] 9301#0: *4558 fastcgi param: "HTTP_COOKIE:
__utmz=206500500.1304337103.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none);
xf_user=1931%2C8b0bad949f0d38b81a46469cb9a941ec05485539;
xf_session=094d80a983afbbd15ac06937d4b23798;
__utma=206500500.1451942054.1304337103.1305433729.1305437755.69;
__utmc=206500500; __utmb=206500500.3.10.1305437755"
2011/05/15 12:43:32 [debug] 9301#0: *4558 posix_memalign: 0894C380:4096
@16
2011/05/15 12:43:32 [debug] 9301#0: *4558 http cleanup add: 0894C400
2011/05/15 12:43:32 [debug] 9301#0: *4558 get rr peer, try: 1
2011/05/15 12:43:32 [debug] 9301#0: *4558 socket 14
2011/05/15 12:43:32 [debug] 9301#0: *4558 epoll add connection: fd:14
ev:80000005
2011/05/15 12:43:32 [debug] 9301#0: *4558 connect to
unix:/var/run/php-fpm.sock, fd:14 #4559
2011/05/15 12:43:32 [debug] 9301#0: *4558 connected
2011/05/15 12:43:32 [debug] 9301#0: *4558 http upstream connect: 0
2011/05/15 12:43:32 [debug] 9301#0: *4558 http upstream send request
2011/05/15 12:43:32 [debug] 9301#0: *4558 chain writer buf fl:0 s:1448
2011/05/15 12:43:32 [debug] 9301#0: *4558 chain writer buf fl:0 s:137
2011/05/15 12:43:32 [debug] 9301#0: *4558 chain writer buf fl:0 s:15
2011/05/15 12:43:32 [debug] 9301#0: *4558 chain writer buf fl:0 s:5
2011/05/15 12:43:32 [debug] 9301#0: *4558 chain writer buf fl:0 s:11
2011/05/15 12:43:32 [debug] 9301#0: *4558 chain writer in: 0894C444
2011/05/15 12:43:32 [debug] 9301#0: *4558 writev: 1616
2011/05/15 12:43:32 [debug] 9301#0: *4558 chain writer out: 00000000
2011/05/15 12:43:32 [debug] 9301#0: *4558 event timer add: 14:
180000:4088501848
2011/05/15 12:43:32 [debug] 9301#0: *4558 http finalize request: -4,
"/index.php?taigachat/list.json" a:1, c:2
2011/05/15 12:43:32 [debug] 9301#0: *4558 http request count:2 blk:0
2011/05/15 12:43:32 [debug] 9301#0: posted event 00000000
2011/05/15 12:43:32 [debug] 9301#0: worker cycle
2011/05/15 12:43:32 [debug] 9301#0: accept mutex locked
2011/05/15 12:43:32 [debug] 9301#0: epoll timer: 198
2011/05/15 12:43:32 [debug] 9301#0: epoll: fd:10 ev:0004 d:08992435
2011/05/15 12:43:32 [debug] 9301#0: *4558 post event 089B80B4
2011/05/15 12:43:32 [debug] 9301#0: epoll: fd:14 ev:0004 d:08992F89
2011/05/15 12:43:32 [debug] 9301#0: *4558 post event 089B8698
2011/05/15 12:43:32 [debug] 9301#0: timer delta: 1
2011/05/15 12:43:32 [debug] 9301#0: posted events 089B8698
2011/05/15 12:43:32 [debug] 9301#0: posted event 089B8698
2011/05/15 12:43:32 [debug] 9301#0: *4558 delete posted event 089B8698
2011/05/15 12:43:32 [debug] 9301#0: *4558 http upstream request:
"/index.php?taigachat/list.json"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http upstream dummy handler
2011/05/15 12:43:32 [debug] 9301#0: posted event 089B80B4
2011/05/15 12:43:32 [debug] 9301#0: *4558 delete posted event 089B80B4
2011/05/15 12:43:32 [debug] 9301#0: *4558 http run request:
"/index.php?taigachat/list.json"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http request empty handler
2011/05/15 12:43:32 [debug] 9301#0: posted event 00000000
2011/05/15 12:43:32 [debug] 9301#0: worker cycle
2011/05/15 12:43:32 [debug] 9301#0: accept mutex locked
2011/05/15 12:43:32 [debug] 9301#0: epoll timer: 197
2011/05/15 12:43:32 [debug] 9301#0: epoll: fd:10 ev:0005 d:08992435
2011/05/15 12:43:32 [debug] 9301#0: *4558 post event 089AB0AC
2011/05/15 12:43:32 [debug] 9301#0: *4558 post event 089B80B4
2011/05/15 12:43:32 [debug] 9301#0: timer delta: 0
2011/05/15 12:43:32 [debug] 9301#0: posted events 089B80B4
2011/05/15 12:43:32 [debug] 9301#0: posted event 089B80B4
2011/05/15 12:43:32 [debug] 9301#0: *4558 delete posted event 089B80B4
2011/05/15 12:43:32 [debug] 9301#0: *4558 http run request:
"/index.php?taigachat/list.json"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http request empty handler
2011/05/15 12:43:32 [debug] 9301#0: posted event 089AB0AC
2011/05/15 12:43:32 [debug] 9301#0: *4558 delete posted event 089AB0AC
2011/05/15 12:43:32 [debug] 9301#0: *4558 http run request:
"/index.php?taigachat/list.json"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http read client request body
2011/05/15 12:43:32 [debug] 9301#0: *4558 add cleanup: 0894C52C
2011/05/15 12:43:32 [debug] 9301#0: *4558 hashed path:
/var/lib/nginx/body/0000000002
2011/05/15 12:43:32 [debug] 9301#0: *4558 temp fd:15
2011/05/15 12:43:32 [warn] 9301#0: *4558 a client request body is
buffered to a temporary file /var/lib/nginx/body/0000000002 while
sending request to upstream, client: [CLIENT IP ADDRESS], server:
example.com, request: "POST /index.php?taigachat/list.json HTTP/1.1",
upstream: "fastcgi://unix:/var/run/php-fpm.sock:", host: "example.com",
referrer: "http://example.com/forum/"
2011/05/15 12:43:32 [debug] 9301#0: *4558 write: 15, 0894B890, 5, 0
2011/05/15 12:43:32 [debug] 9301#0: *4558 recv: fd:10 0 of 0
2011/05/15 12:43:32 [debug] 9301#0: *4558 http client request body recv
0
2011/05/15 12:43:32 [info] 9301#0: *4558 client closed prematurely
connection while sending request to upstream, client: [CLIENT IP
ADDRESS], server: example.com, request: "POST
/index.php?taigachat/list.json HTTP/1.1", upstream:
"fastcgi://unix:/var/run/php-fpm.sock:", host: "example.com", referrer:
"http://example.com/forum/"
2011/05/15 12:43:32 [debug] 9301#0: *4558 http finalize request: 400,
"/index.php?taigachat/list.json" a:1, c:1
2011/05/15 12:43:32 [debug] 9301#0: *4558 http terminate request
count:1
2011/05/15 12:43:32 [debug] 9301#0: *4558 cleanup http upstream request:
"/index.php"
2011/05/15 12:43:32 [debug] 9301#0: *4558 finalize http upstream
request: -4
2011/05/15 12:43:32 [debug] 9301#0: *4558 finalize http fastcgi request
2011/05/15 12:43:32 [debug] 9301#0: *4558 free rr peer 1 0
2011/05/15 12:43:32 [debug] 9301#0: *4558 close http upstream
connection: 14
2011/05/15 12:43:32 [debug] 9301#0: *4558 event timer del: 14:
4088501848
2011/05/15 12:43:32 [debug] 9301#0: *4558 reusable connection: 0
2011/05/15 12:43:32 [debug] 9301#0: *4558 http finalize request: -4,
"/index.php?taigachat/list.json" a:1, c:1
2011/05/15 12:43:32 [debug] 9301#0: *4558 set http keepalive handler
2011/05/15 12:43:32 [debug] 9301#0: *4558 http close request
2011/05/15 12:43:32 [debug] 9301#0: *4558 http log handler
2011/05/15 12:43:32 [debug] 9301#0: *4558 run cleanup: 0894C52C
2011/05/15 12:43:32 [debug] 9301#0: *4558 file cleanup: fd:15
2011/05/15 12:43:32 [debug] 9301#0: *4558 free: 0894AF70, unused: 5
2011/05/15 12:43:32 [debug] 9301#0: *4558 free: 0894C380, unused: 3417
2011/05/15 12:43:32 [debug] 9301#0: *4558 event timer add: 10:
5000:4088326849
2011/05/15 12:43:32 [debug] 9301#0: *4558 free: 08985548
2011/05/15 12:43:32 [debug] 9301#0: *4558 free: 08983960
2011/05/15 12:43:32 [debug] 9301#0: *4558 hc free: 00000000 0
2011/05/15 12:43:32 [debug] 9301#0: *4558 hc busy: 00000000 0
2011/05/15 12:43:32 [debug] 9301#0: *4558 tcp_nodelay
2011/05/15 12:43:32 [debug] 9301#0: *4558 reusable connection: 1
2011/05/15 12:43:32 [notice] 9299#0: signal 17 (SIGCHLD) received
2011/05/15 12:43:32 [alert] 9299#0: worker process 9301 exited on signal
11 (core dumped)
2011/05/15 12:43:32 [debug] 9299#0: wake up, sigio 0
2011/05/15 12:43:32 [debug] 9299#0: reap children
2011/05/15 12:43:32 [debug] 9299#0: child: 0 9812 e:0 t:0 d:0 r:1 j:0
2011/05/15 12:43:32 [debug] 9299#0: child: 1 9301 e:0 t:1 d:0 r:1 j:0
2011/05/15 12:43:32 [debug] 9299#0: pass close channel s:1 pid:9301
to:9812
2011/05/15 12:43:32 [debug] 9299#0: channel 10:11
2011/05/15 12:43:32 [debug] 9812#0: epoll: fd:9 ev:0001 d:08991D2C
2011/05/15 12:43:32 [debug] 9812#0: channel handler
2011/05/15 12:43:32 [debug] 9812#0: channel: 16
2011/05/15 12:43:32 [debug] 9812#0: channel command: 2
2011/05/15 12:43:32 [debug] 9812#0: close channel s:1 pid:9301 our:9301
fd:10
2011/05/15 12:43:32 [debug] 9812#0: channel: -2
2011/05/15 12:43:32 [debug] 9812#0: timer delta: 112
2011/05/15 12:43:32 [debug] 9812#0: posted events 00000000
2011/05/15 12:43:32 [debug] 9812#0: worker cycle
2011/05/15 12:43:32 [debug] 9812#0: accept mutex locked
2011/05/15 12:43:32 [debug] 9812#0: epoll add event: fd:8 op:1
ev:00000001
2011/05/15 12:43:32 [debug] 9812#0: epoll timer: -1
2011/05/15 12:43:32 [notice] 9299#0: start worker process 9813
2011/05/15 12:43:32 [debug] 9299#0: pass channel s:1 pid:9813 fd:10 to
s:0 pid:9812 fd:3
2011/05/15 12:43:32 [debug] 9299#0: sigsuspend
2011/05/15 12:43:32 [debug] 9812#0: epoll: fd:9 ev:0001 d:08991D2C
2011/05/15 12:43:32 [debug] 9812#0: post event 089AAD04
2011/05/15 12:43:32 [debug] 9812#0: timer delta: 1
2011/05/15 12:43:32 [notice] 9299#0: signal 29 (SIGIO) received
2011/05/15 12:43:32 [debug] 9812#0: posted events 089AAD04
2011/05/15 12:43:32 [debug] 9299#0: wake up, sigio 0
2011/05/15 12:43:32 [debug] 9812#0: posted event 089AAD04
2011/05/15 12:43:32 [debug] 9299#0: sigsuspend
2011/05/15 12:43:32 [debug] 9812#0: delete posted event 089AAD04
2011/05/15 12:43:32 [debug] 9812#0: channel handler
2011/05/15 12:43:32 [debug] 9812#0: channel: 16
2011/05/15 12:43:32 [debug] 9812#0: channel command: 1
2011/05/15 12:43:32 [debug] 9812#0: get channel s:1 pid:9813 fd:3
2011/05/15 12:43:32 [debug] 9812#0: channel: -2
2011/05/15 12:43:32 [debug] 9812#0: posted event 00000000
2011/05/15 12:43:32 [debug] 9812#0: worker cycle
2011/05/15 12:43:32 [debug] 9812#0: accept mutex locked
2011/05/15 12:43:32 [debug] 9812#0: epoll timer: -1
2011/05/15 12:43:32 [notice] 9813#0: sched_setaffinity(0x0000000A)
2011/05/15 12:43:32 [debug] 9813#0: malloc: 0895AAA8:6144
2011/05/15 12:43:32 [debug] 9813#0: malloc: 08991CC8:102400
2011/05/15 12:43:32 [debug] 9813#0: malloc: 089AACD0:53248
2011/05/15 12:43:32 [debug] 9813#0: malloc: 089B7CD8:53248
2011/05/15 12:43:32 [debug] 9813#0: epoll add event: fd:11 op:1
ev:00000001
2011/05/15 12:43:32 [debug] 9813#0: setproctitle: "nginx: worker
process"
2011/05/15 12:43:32 [debug] 9813#0: worker cycle
2011/05/15 12:43:32 [debug] 9813#0: accept mutex lock failed: 0
2011/05/15 12:43:32 [debug] 9813#0: epoll timer: 500

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198003,198228#msg-198228



From nginx-forum at nginx.us  Sun May 15 18:49:34 2011
From: nginx-forum at nginx.us (holstebm)
Date: Sun, 15 May 2011 10:49:34 -0400
Subject: add_header works in server but not in http?
In-Reply-To: <20110515125221.GE79835@sysoev.ru>
References: <20110515125221.GE79835@sysoev.ru>
Message-ID: <e3b95321490ebbd7790b3ba0c67fcad9.NginxMailingListEnglish@forum.nginx.org>

I feel a little foolish.  I checked the WIKI and it says "Note that it
just appends a new header entry to the output header list." but doesn't
say anything about the behavior you describe.  I guess I hadn't noted
this until I started moving things around so that they make more logical
sense in my config files.  I had thought add_header to be cumulative
from parent to child when the headers were different (i.e. Cache-Control
in one block, X-UA-Compatible in another).

I didn't mean to waste your time with this.  I know this is a huge
project and I very much appreciate your hard work.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198143,198231#msg-198231



From nginx-forum at nginx.us  Sun May 15 19:44:17 2011
From: nginx-forum at nginx.us (inph)
Date: Sun, 15 May 2011 11:44:17 -0400
Subject: nginx 1.0.2 intermittent segfault error 4
In-Reply-To: <4c5ea4c84247919300aa51e1226efb52.NginxMailingListEnglish@forum.nginx.org>
References: <041aa0fb00af03026d71bcb288c84db7.NginxMailingListEnglish@forum.nginx.org>
	<4c5ea4c84247919300aa51e1226efb52.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <0ea59159f557484bc42c72eb1ff8f868.NginxMailingListEnglish@forum.nginx.org>

During low day time traffic, less than 3 hours after the first segfault
above. Same parent nginx process.

[93995.665429] nginx[9301]: segfault at 50545448 ip 08085243 sp bfdb0b50
error 4 in nginx[8048000+97000]
[103616.200896] nginx[9813]: segfault at 50545448 ip 08085243 sp
bfdb0b30 error 4 in nginx[8048000+97000]

# stat core
  File: `core'
  Size: 1515520         Blocks: 2416       IO Block: 4096   regular
file
Device: 807h/2055d      Inode: 18350098    Links: 1
Access: (0600/-rw-------)  Uid: (   48/     www)   Gid: (   48/    
www)
Access: 2011-05-15 15:23:53.000000000 +0000
Modify: 2011-05-15 15:23:53.000000000 +0000
Change: 2011-05-15 15:23:53.000000000 +0000
 Birth: -

# nginx -V
nginx: nginx version: nginx/1.0.2
nginx: TLS SNI support enabled
nginx: configure arguments: --with-debug --prefix=/usr
--sbin-path=/usr/sbin/nginx --conf-path=/etc/nginx/nginx.conf
--pid-path=/var/run/nginx.pid --lock-path=/var/lock/nginx.lock
--user=www --group=www --error-log-path=/var/log/nginx/error.log
--http-log-path=/var/log/nginx/access.log --with-http_ssl_module
--with-http_realip_module --with-http_addition_module
--with-http_sub_module --with-http_dav_module --with-http_flv_module
--with-http_gzip_static_module --with-http_secure_link_module
--with-http_stub_status_module
--http-client-body-temp-path=/var/lib/nginx/body
--http-proxy-temp-path=/var/lib/nginx/proxy
--http-fastcgi-temp-path=/var/lib/nginx/fastcgi
--http-uwsgi-temp-path=/var/lib/nginx/uwsgi
--http-scgi-temp-path=/var/lib/nginx/scgi --without-mail_pop3_module
--without-mail_imap_module --without-mail_smtp_module

# gdb /usr/sbin/nginx /home/www/core
GNU gdb (GDB) 7.2
Copyright (C) 2010 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later
<http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show
copying"
and "show warranty" for details.
This GDB was configured as "i486-slackware-linux".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/sbin/nginx...done.
[New Thread 9813]

warning: Can't read pathname for load map: Input/output error.
[snip no symbols found in libs, identical to previous posts]
Core was generated by `nginx:'.
Program terminated with signal 11, Segmentation fault.
#0  0x08085243 in ngx_http_terminate_request (r=0x894d308, rc=400) at
src/http/ngx_http_request.c:2085
2085        ngx_log_debug2(NGX_LOG_DEBUG_HTTP, r->connection->log, 0,
(gdb) bt
#0  0x08085243 in ngx_http_terminate_request (r=0x894d308, rc=400) at
src/http/ngx_http_request.c:2085
#1  0x08084c91 in ngx_http_finalize_request (r=0x894d308, rc=400) at
src/http/ngx_http_request.c:1912
#2  0x0808dab7 in ngx_http_read_client_request_body_handler
(r=0x894d308) at src/http/ngx_http_request_body.c:254
#3  0x080849ab in ngx_http_request_handler (ev=0x89aad38) at
src/http/ngx_http_request.c:1800
#4  0x08067f7b in ngx_event_process_posted (cycle=0x8946cb8,
posted=0x80eaaf4) at src/event/ngx_event_posted.c:39
#5  0x08066531 in ngx_process_events_and_timers (cycle=0x8946cb8) at
src/event/ngx_event.c:272
#6  0x0807051b in ngx_worker_process_cycle (cycle=0x8946cb8, data=0x0)
at src/os/unix/ngx_process_cycle.c:800
#7  0x0806dc5f in ngx_spawn_process (cycle=0x8946cb8, proc=0x807038a
<ngx_worker_process_cycle>, data=0x0, name=0x80d2e2b "worker process",
respawn=1)
    at src/os/unix/ngx_process.c:196
#8  0x080700bd in ngx_reap_children (cycle=0x8946cb8) at
src/os/unix/ngx_process_cycle.c:617
#9  0x0806f26b in ngx_master_process_cycle (cycle=0x8946cb8) at
src/os/unix/ngx_process_cycle.c:180
#10 0x0804c297 in main (argc=1, argv=0xbfdb0fc4) at
src/core/nginx.c:405
(gdb) bt full
#0  0x08085243 in ngx_http_terminate_request (r=0x894d308, rc=400) at
src/http/ngx_http_request.c:2085
        cln = 0x0
        mr = 0x894d308
        e = 0x89850e8
#1  0x08084c91 in ngx_http_finalize_request (r=0x894d308, rc=400) at
src/http/ngx_http_request.c:1912
        c = 0x8991d90
        pr = 0x8991d90
        clcf = 0x894deb4
#2  0x0808dab7 in ngx_http_read_client_request_body_handler
(r=0x894d308) at src/http/ngx_http_request_body.c:254
        rc = 400
#3  0x080849ab in ngx_http_request_handler (ev=0x89aad38) at
src/http/ngx_http_request.c:1800
        c = 0x8991d90
        r = 0x894d308
        ctx = 0x8985110
#4  0x08067f7b in ngx_event_process_posted (cycle=0x8946cb8,
posted=0x80eaaf4) at src/event/ngx_event_posted.c:39
        ev = 0x89aad38
#5  0x08066531 in ngx_process_events_and_timers (cycle=0x8946cb8) at
src/event/ngx_event.c:272
        flags = 3
        timer = 179999
        delta = 1
#6  0x0807051b in ngx_worker_process_cycle (cycle=0x8946cb8, data=0x0)
at src/os/unix/ngx_process_cycle.c:800
        i = 9299
        c = 0xbfdb0c98
#7  0x0806dc5f in ngx_spawn_process (cycle=0x8946cb8, proc=0x807038a
<ngx_worker_process_cycle>, data=0x0, name=0x80d2e2b "worker process",
respawn=1)
    at src/os/unix/ngx_process.c:196
        on = 1
        pid = 0
        s = 1
#8  0x080700bd in ngx_reap_children (cycle=0x8946cb8) at
src/os/unix/ngx_process_cycle.c:617
        i = 1
        n = 2
        live = 1
        ch = {command = 2, pid = 9301, slot = 1, fd = -1}
        ccf = 0xbfdb0d10
#9  0x0806f26b in ngx_master_process_cycle (cycle=0x8946cb8) at
src/os/unix/ngx_process_cycle.c:180
        title = 0x898feb6 "master process nginx"
        p = 0x898feca ""
        size = 21
        i = 1
        n = 143944900
        sigio = 0
        set = {__val = {0 <repeats 32 times>}}
        itv = {it_interval = {tv_sec = 135072442, tv_usec = 9299},
it_value = {tv_sec = 0, tv_usec = 0}}
        live = 1
        delay = 0
        ls = 0x0
        ccf = 0x89474b0
#10 0x0804c297 in main (argc=1, argv=0xbfdb0fc4) at
src/core/nginx.c:405
        i = 53
        log = 0x80e88d0
        cycle = 0x8946cb8
        init_cycle = {conf_ctx = 0x0, pool = 0x8945dd0, log = 0x80e88d0,
new_log = {log_level = 0, file = 0x0, connection = 0, handler = 0, data
= 0x0, action = 0x0},
          files = 0x0, free_connections = 0x0, free_connection_n = 0,
reusable_connections_queue = {prev = 0x0, next = 0x0}, listening = {elts
= 0x0, nelts = 0, size = 0,
            nalloc = 0, pool = 0x0}, pathes = {elts = 0x0, nelts = 0,
size = 0, nalloc = 0, pool = 0x0}, open_files = {last = 0x0, part =
{elts = 0x0, nelts = 0, next = 0x0},
            size = 0, nalloc = 0, pool = 0x0}, shared_memory = {last =
0x0, part = {elts = 0x0, nelts = 0, next = 0x0}, size = 0, nalloc = 0,
pool = 0x0}, connection_n = 0,
          files_n = 0, connections = 0x0, read_events = 0x0,
write_events = 0x0, old_cycle = 0x0, conf_file = {len = 21, data =
0x80cf5c2 "/etc/nginx/nginx.conf"},
          conf_param = {len = 0, data = 0x0}, conf_prefix = {len = 11,
data = 0x80cf5c2 "/etc/nginx/nginx.conf"}, prefix = {len = 5, data =
0x80cf5bc "/usr/"}, lock_file = {
            len = 0, data = 0x0}, hostname = {len = 0, data = 0x0}}
        ccf = 0x89474b0
(gdb) quit


# cat segfault.log |grep "15:23:53"
2011/05/15 15:23:53 [debug] 9813#0: epoll: fd:10 ev:0001 d:08991D91
2011/05/15 15:23:53 [debug] 9813#0: *7244 post event 089AAD38
2011/05/15 15:23:53 [debug] 9813#0: timer delta: 335
2011/05/15 15:23:53 [debug] 9813#0: posted events 089AAD38
2011/05/15 15:23:53 [debug] 9813#0: posted event 089AAD38
2011/05/15 15:23:53 [debug] 9813#0: *7244 delete posted event 089AAD38
2011/05/15 15:23:53 [debug] 9813#0: *7244 http process request header
line
2011/05/15 15:23:53 [debug] 9813#0: *7244 recv: fd:10 488 of 488
2011/05/15 15:23:53 [debug] 9813#0: *7244 http header: "Accept-Charset:
ISO-8859-1,utf-8;q=0.7,*;q=0.3"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http header: "Cookie:
__utmz=206500500.1304337103.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none);
xf_user=1931%2C8b0bad949f0d38b81a46469cb9a941ec05485539;
xf_session=9e29ef01aa56202856bbf74cbae5aed7;
__utma=206500500.1451942054.1304337103.1305442622.1305444878.71;
__utmc=206500500; __utmb=206500500.7.10.1305444878"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http header done
2011/05/15 15:23:53 [debug] 9813#0: *7244 event timer del: 10:
4097999971
2011/05/15 15:23:53 [debug] 9813#0: *7244 generic phase: 0
2011/05/15 15:23:53 [debug] 9813#0: *7244 rewrite phase: 1
2011/05/15 15:23:53 [debug] 9813#0: *7244 test location: "/"
2011/05/15 15:23:53 [debug] 9813#0: *7244 test location: "forum/"
2011/05/15 15:23:53 [debug] 9813#0: *7244 test location: ~
"\.(jpg|jpeg|gif|css|png|js|ico)$"
2011/05/15 15:23:53 [debug] 9813#0: *7244 test location: ~ "\.php$"
2011/05/15 15:23:53 [debug] 9813#0: *7244 using configuration "\.php$"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http cl:142 max:52428800
2011/05/15 15:23:53 [debug] 9813#0: *7244 rewrite phase: 3
2011/05/15 15:23:53 [debug] 9813#0: *7244 post rewrite phase: 4
2011/05/15 15:23:53 [debug] 9813#0: *7244 generic phase: 5
2011/05/15 15:23:53 [debug] 9813#0: *7244 generic phase: 6
2011/05/15 15:23:53 [debug] 9813#0: *7244 generic phase: 7
2011/05/15 15:23:53 [debug] 9813#0: *7244 access phase: 8
2011/05/15 15:23:53 [debug] 9813#0: *7244 access phase: 9
2011/05/15 15:23:53 [debug] 9813#0: *7244 post access phase: 10
2011/05/15 15:23:53 [debug] 9813#0: *7244 try files phase: 11
2011/05/15 15:23:53 [debug] 9813#0: *7244 http client request body
preread 137
2011/05/15 15:23:53 [debug] 9813#0: *7244 http read client request body
2011/05/15 15:23:53 [debug] 9813#0: *7244 recv: fd:10 5 of 5
2011/05/15 15:23:53 [debug] 9813#0: *7244 http client request body recv
5
2011/05/15 15:23:53 [debug] 9813#0: *7244 http client request body rest
0
2011/05/15 15:23:53 [debug] 9813#0: *7244 http init upstream, client
timer: 0
2011/05/15 15:23:53 [debug] 9813#0: *7244 epoll add event: fd:10 op:3
ev:80000005
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"SCRIPT_FILENAME"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var:
"/home/www/example.com/htdocs"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var: "/index.php"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"SCRIPT_FILENAME: /home/www/example.com/htdocs/index.php"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"QUERY_STRING"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var:
"taigachat/list.json"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "QUERY_STRING:
taigachat/list.json"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"REQUEST_METHOD"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var: "POST"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"REQUEST_METHOD: POST"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"CONTENT_TYPE"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var:
"application/x-www-form-urlencoded"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "CONTENT_TYPE:
application/x-www-form-urlencoded"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"CONTENT_LENGTH"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var: "142"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"CONTENT_LENGTH: 142"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"SCRIPT_NAME"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var: "/index.php"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "SCRIPT_NAME:
/index.php"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"REQUEST_URI"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var:
"/index.php?taigachat/list.json"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "REQUEST_URI:
/index.php?taigachat/list.json"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"DOCUMENT_URI"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var: "/index.php"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "DOCUMENT_URI:
/index.php"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"DOCUMENT_ROOT"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var:
"/home/www/example.com/htdocs"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "DOCUMENT_ROOT:
/home/www/example.com/htdocs"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"SERVER_PROTOCOL"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var: "HTTP/1.1"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"SERVER_PROTOCOL: HTTP/1.1"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"GATEWAY_INTERFACE"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy: "CGI/1.1"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"GATEWAY_INTERFACE: CGI/1.1"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"SERVER_SOFTWARE"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy: "nginx/"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var: "1.0.2"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"SERVER_SOFTWARE: nginx/1.0.2"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"REMOTE_ADDR"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var: "[CLIENT IP
ADDRESS]"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "REMOTE_ADDR:
[CLIENT IP ADDRESS]"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"REMOTE_PORT"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var: "52898"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "REMOTE_PORT:
52898"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"SERVER_ADDR"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var: "[SERVER IP
ADDRESS]"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "SERVER_ADDR:
[SERVER IP ADDRESS]"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"SERVER_PORT"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var: "80"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "SERVER_PORT:
80"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"SERVER_NAME"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script var:
"example.com"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "SERVER_NAME:
example.com"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy:
"REDIRECT_STATUS"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http script copy: "200"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"REDIRECT_STATUS: 200"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "HTTP_HOST:
example.com"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"HTTP_CONNECTION: keep-alive"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "HTTP_REFERER:
http://example.com/forum/"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"HTTP_CONTENT_LENGTH: 142"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "HTTP_ORIGIN:
http://example.com"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"HTTP_X_REQUESTED_WITH: XMLHttpRequest"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"HTTP_USER_AGENT: Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/534.24
(KHTML, like Gecko) Chrome/11.0.696.65 Safari/534.24"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"HTTP_CONTENT_TYPE: application/x-www-form-urlencoded"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "HTTP_ACCEPT:
application/json, text/javascript, */*; q=0.01"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"HTTP_ACCEPT_ENCODING: gzip,deflate"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"HTTP_ACCEPT_LANGUAGE: da-DK,da;q=0.8,en-US;q=0.6,en;q=0.4"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param:
"HTTP_ACCEPT_CHARSET: ISO-8859-1,utf-8;q=0.7,*;q=0.3"
2011/05/15 15:23:53 [debug] 9813#0: *7244 fastcgi param: "HTTP_COOKIE:
__utmz=206500500.1304337103.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none);
xf_user=1931%2C8b0bad949f0d38b81a46469cb9a941ec05485539;
xf_session=9e29ef01aa56202856bbf74cbae5aed7;
__utma=206500500.1451942054.1304337103.1305442622.1305444878.71;
__utmc=206500500; __utmb=206500500.7.10.1305444878"
2011/05/15 15:23:53 [debug] 9813#0: *7244 posix_memalign: 0894E640:4096
@16
2011/05/15 15:23:53 [debug] 9813#0: *7244 http cleanup add: 0894E6C0
2011/05/15 15:23:53 [debug] 9813#0: *7244 get rr peer, try: 1
2011/05/15 15:23:53 [debug] 9813#0: *7244 socket 7
2011/05/15 15:23:53 [debug] 9813#0: *7244 epoll add connection: fd:7
ev:80000005
2011/05/15 15:23:53 [debug] 9813#0: *7244 connect to
unix:/var/run/php-fpm.sock, fd:7 #7245
2011/05/15 15:23:53 [debug] 9813#0: *7244 connected
2011/05/15 15:23:53 [debug] 9813#0: *7244 http upstream connect: 0
2011/05/15 15:23:53 [debug] 9813#0: *7244 http upstream send request
2011/05/15 15:23:53 [debug] 9813#0: *7244 chain writer buf fl:0 s:1448
2011/05/15 15:23:53 [debug] 9813#0: *7244 chain writer buf fl:0 s:137
2011/05/15 15:23:53 [debug] 9813#0: *7244 chain writer buf fl:0 s:15
2011/05/15 15:23:53 [debug] 9813#0: *7244 chain writer buf fl:0 s:5
2011/05/15 15:23:53 [debug] 9813#0: *7244 chain writer buf fl:0 s:11
2011/05/15 15:23:53 [debug] 9813#0: *7244 chain writer in: 0894E704
2011/05/15 15:23:53 [debug] 9813#0: *7244 writev: 1616
2011/05/15 15:23:53 [debug] 9813#0: *7244 chain writer out: 00000000
2011/05/15 15:23:53 [debug] 9813#0: *7244 event timer add: 7:
180000:4098122382
2011/05/15 15:23:53 [debug] 9813#0: *7244 http finalize request: -4,
"/index.php?taigachat/list.json" a:1, c:2
2011/05/15 15:23:53 [debug] 9813#0: *7244 http request count:2 blk:0
2011/05/15 15:23:53 [debug] 9813#0: posted event 00000000
2011/05/15 15:23:53 [debug] 9813#0: worker cycle
2011/05/15 15:23:53 [debug] 9813#0: accept mutex locked
2011/05/15 15:23:53 [debug] 9813#0: epoll timer: 180000
2011/05/15 15:23:53 [debug] 9813#0: epoll: fd:10 ev:0004 d:08991D91
2011/05/15 15:23:53 [debug] 9813#0: *7244 post event 089B7D40
2011/05/15 15:23:53 [debug] 9813#0: epoll: fd:7 ev:0004 d:089920B1
2011/05/15 15:23:53 [debug] 9813#0: *7244 post event 089B7EE0
2011/05/15 15:23:53 [debug] 9813#0: timer delta: 1
2011/05/15 15:23:53 [debug] 9813#0: posted events 089B7EE0
2011/05/15 15:23:53 [debug] 9813#0: posted event 089B7EE0
2011/05/15 15:23:53 [debug] 9813#0: *7244 delete posted event 089B7EE0
2011/05/15 15:23:53 [debug] 9813#0: *7244 http upstream request:
"/index.php?taigachat/list.json"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http upstream dummy handler
2011/05/15 15:23:53 [debug] 9813#0: posted event 089B7D40
2011/05/15 15:23:53 [debug] 9813#0: *7244 delete posted event 089B7D40
2011/05/15 15:23:53 [debug] 9813#0: *7244 http run request:
"/index.php?taigachat/list.json"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http request empty handler
2011/05/15 15:23:53 [debug] 9813#0: posted event 00000000
2011/05/15 15:23:53 [debug] 9813#0: worker cycle
2011/05/15 15:23:53 [debug] 9813#0: accept mutex locked
2011/05/15 15:23:53 [debug] 9813#0: epoll timer: 179999
2011/05/15 15:23:53 [debug] 9813#0: epoll: fd:10 ev:0005 d:08991D91
2011/05/15 15:23:53 [debug] 9813#0: *7244 post event 089AAD38
2011/05/15 15:23:53 [debug] 9813#0: *7244 post event 089B7D40
2011/05/15 15:23:53 [debug] 9813#0: timer delta: 1
2011/05/15 15:23:53 [debug] 9813#0: posted events 089B7D40
2011/05/15 15:23:53 [debug] 9813#0: posted event 089B7D40
2011/05/15 15:23:53 [debug] 9813#0: *7244 delete posted event 089B7D40
2011/05/15 15:23:53 [debug] 9813#0: *7244 http run request:
"/index.php?taigachat/list.json"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http request empty handler
2011/05/15 15:23:53 [debug] 9813#0: posted event 089AAD38
2011/05/15 15:23:53 [debug] 9813#0: *7244 delete posted event 089AAD38
2011/05/15 15:23:53 [debug] 9813#0: *7244 http run request:
"/index.php?taigachat/list.json"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http read client request body
2011/05/15 15:23:53 [debug] 9813#0: *7244 add cleanup: 0894E7EC
2011/05/15 15:23:53 [debug] 9813#0: *7244 hashed path:
/var/lib/nginx/body/0000000004
2011/05/15 15:23:53 [debug] 9813#0: *7244 temp fd:14
2011/05/15 15:23:53 [warn] 9813#0: *7244 a client request body is
buffered to a temporary file /var/lib/nginx/body/0000000004 while
sending request to upstream, client: [CLIENT IP ADDRESS], server:
example.com, request: "POST /index.php?taigachat/list.json HTTP/1.1",
upstream: "fastcgi://unix:/var/run/php-fpm.sock:", host: "example.com",
referrer: "http://example.com/forum/"
2011/05/15 15:23:53 [debug] 9813#0: *7244 write: 14, 0894DF40, 5, 0
2011/05/15 15:23:53 [debug] 9813#0: *7244 recv: fd:10 0 of 0
2011/05/15 15:23:53 [debug] 9813#0: *7244 http client request body recv
0
2011/05/15 15:23:53 [info] 9813#0: *7244 client closed prematurely
connection while sending request to upstream, client: [CLIENT IP
ADDRESS], server: example.com, request: "POST
/index.php?taigachat/list.json HTTP/1.1", upstream:
"fastcgi://unix:/var/run/php-fpm.sock:", host: "example.com", referrer:
"http://example.com/forum/"
2011/05/15 15:23:53 [debug] 9813#0: *7244 http finalize request: 400,
"/index.php?taigachat/list.json" a:1, c:1
2011/05/15 15:23:53 [debug] 9813#0: *7244 http terminate request
count:1
2011/05/15 15:23:53 [debug] 9813#0: *7244 cleanup http upstream request:
"/index.php"
2011/05/15 15:23:53 [debug] 9813#0: *7244 finalize http upstream
request: -4
2011/05/15 15:23:53 [debug] 9813#0: *7244 finalize http fastcgi request
2011/05/15 15:23:53 [debug] 9813#0: *7244 free rr peer 1 0
2011/05/15 15:23:53 [debug] 9813#0: *7244 close http upstream
connection: 7
2011/05/15 15:23:53 [debug] 9813#0: *7244 event timer del: 7:
4098122382
2011/05/15 15:23:53 [debug] 9813#0: *7244 reusable connection: 0
2011/05/15 15:23:53 [debug] 9813#0: *7244 http finalize request: -4,
"/index.php?taigachat/list.json" a:1, c:1
2011/05/15 15:23:53 [debug] 9813#0: *7244 set http keepalive handler
2011/05/15 15:23:53 [debug] 9813#0: *7244 http close request
2011/05/15 15:23:53 [debug] 9813#0: *7244 http log handler
2011/05/15 15:23:53 [debug] 9813#0: *7244 run cleanup: 0894E7EC
2011/05/15 15:23:53 [debug] 9813#0: *7244 file cleanup: fd:14
2011/05/15 15:23:53 [debug] 9813#0: *7244 free: 0894D620, unused: 5
2011/05/15 15:23:53 [debug] 9813#0: *7244 free: 0894E640, unused: 3417
2011/05/15 15:23:53 [debug] 9813#0: *7244 event timer add: 10:
5000:4097947384
2011/05/15 15:23:53 [debug] 9813#0: *7244 free: 0894D308
2011/05/15 15:23:53 [debug] 9813#0: *7244 free: 089855B8
2011/05/15 15:23:53 [debug] 9813#0: *7244 hc free: 00000000 0
2011/05/15 15:23:53 [debug] 9813#0: *7244 hc busy: 00000000 0
2011/05/15 15:23:53 [debug] 9813#0: *7244 tcp_nodelay
2011/05/15 15:23:53 [debug] 9813#0: *7244 reusable connection: 1
2011/05/15 15:23:53 [notice] 9299#0: signal 17 (SIGCHLD) received
2011/05/15 15:23:53 [alert] 9299#0: worker process 9813 exited on signal
11 (core dumped)
2011/05/15 15:23:53 [debug] 9299#0: wake up, sigio 0
2011/05/15 15:23:53 [debug] 9299#0: reap children
2011/05/15 15:23:53 [debug] 9299#0: child: 0 9812 e:0 t:0 d:0 r:1 j:0
2011/05/15 15:23:53 [debug] 9299#0: child: 1 9813 e:0 t:1 d:0 r:1 j:0
2011/05/15 15:23:53 [debug] 9299#0: pass close channel s:1 pid:9813
to:9812
2011/05/15 15:23:53 [debug] 9299#0: channel 7:10
2011/05/15 15:23:53 [notice] 9299#0: start worker process 10069
2011/05/15 15:23:53 [debug] 9299#0: pass channel s:1 pid:10069 fd:7 to
s:0 pid:9812 fd:3
2011/05/15 15:23:53 [debug] 9299#0: sigsuspend
2011/05/15 15:23:53 [debug] 9812#0: epoll: fd:9 ev:0001 d:08991D2C
2011/05/15 15:23:53 [debug] 9812#0: channel handler
2011/05/15 15:23:53 [debug] 9812#0: channel: 16
2011/05/15 15:23:53 [debug] 9812#0: channel command: 2
2011/05/15 15:23:53 [debug] 9812#0: close channel s:1 pid:9813 our:9813
fd:3
2011/05/15 15:23:53 [debug] 9812#0: channel: 16
2011/05/15 15:23:53 [debug] 9812#0: channel command: 1
2011/05/15 15:23:53 [debug] 9812#0: get channel s:1 pid:10069 fd:3
2011/05/15 15:23:53 [debug] 9812#0: channel: -2
2011/05/15 15:23:53 [debug] 9812#0: timer delta: 243
2011/05/15 15:23:53 [debug] 9812#0: posted events 00000000
2011/05/15 15:23:53 [debug] 9812#0: worker cycle
2011/05/15 15:23:53 [debug] 9812#0: accept mutex locked
2011/05/15 15:23:53 [debug] 9812#0: epoll add event: fd:8 op:1
ev:00000001
2011/05/15 15:23:53 [debug] 9812#0: epoll timer: -1
2011/05/15 15:23:53 [notice] 9299#0: signal 29 (SIGIO) received
2011/05/15 15:23:53 [debug] 9299#0: wake up, sigio 0
2011/05/15 15:23:53 [debug] 9299#0: sigsuspend
2011/05/15 15:23:53 [notice] 10069#0: sched_setaffinity(0x0000000A)
2011/05/15 15:23:53 [debug] 10069#0: malloc: 0895AAA8:6144
2011/05/15 15:23:53 [debug] 10069#0: malloc: 08991CC8:102400
2011/05/15 15:23:53 [debug] 10069#0: malloc: 089AACD0:53248
2011/05/15 15:23:53 [debug] 10069#0: malloc: 089B7CD8:53248
2011/05/15 15:23:53 [debug] 10069#0: epoll add event: fd:10 op:1
ev:00000001
2011/05/15 15:23:53 [debug] 10069#0: setproctitle: "nginx: worker
process"
2011/05/15 15:23:53 [debug] 10069#0: worker cycle
2011/05/15 15:23:53 [debug] 10069#0: accept mutex lock failed: 0
2011/05/15 15:23:53 [debug] 10069#0: epoll timer: 500
2011/05/15 15:23:53 [debug] 10069#0: timer delta: 503
2011/05/15 15:23:53 [debug] 10069#0: posted events 00000000
2011/05/15 15:23:53 [debug] 10069#0: worker cycle
2011/05/15 15:23:53 [debug] 10069#0: accept mutex lock failed: 0
2011/05/15 15:23:53 [debug] 10069#0: epoll timer: 500

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198003,198243#msg-198243



From rschultz7 at gmail.com  Sun May 15 20:10:53 2011
From: rschultz7 at gmail.com (Rob Schultz)
Date: Sun, 15 May 2011 18:10:53 +0200
Subject: IPv6 issues viewing any static or dynamic file
In-Reply-To: <20110515122627.GA79835@sysoev.ru>
References: <20024305-77F2-45ED-9D1D-D189407A1C0E@gmail.com>
	<20110515122627.GA79835@sysoev.ru>
Message-ID: <D9A64039-7C66-4E93-BA5F-8714319B4F2B@gmail.com>


On May 15, 2011, at 2:26 PM, Igor Sysoev wrote:

> 
> How do you connect via IPv6 ?
> It seems like large packets tunneling issue.
> Are you able to scp large file via IPv6 ?

I am using a 6to4 HE tunnel on the client side and native IPv6 on the server side. (It's hosted at Linode that recently enabled native IPv6) 

You are right. I created just a test.txt file with just hello world in it and i can browse it just fine. I wonder what google is doing because i can browse ipv6.google.com just fine and other test sights. 

Just modified my local MTU to 1280 and it seems to be working now. Sorry for the noise. 

v/r,
Rob

From igor at sysoev.ru  Sun May 15 20:23:16 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Sun, 15 May 2011 20:23:16 +0400
Subject: IPv6 issues viewing any static or dynamic file
In-Reply-To: <D9A64039-7C66-4E93-BA5F-8714319B4F2B@gmail.com>
References: <20024305-77F2-45ED-9D1D-D189407A1C0E@gmail.com>
	<20110515122627.GA79835@sysoev.ru>
	<D9A64039-7C66-4E93-BA5F-8714319B4F2B@gmail.com>
Message-ID: <20110515162315.GA96402@sysoev.ru>

On Sun, May 15, 2011 at 06:10:53PM +0200, Rob Schultz wrote:
> 
> On May 15, 2011, at 2:26 PM, Igor Sysoev wrote:
> 
> > 
> > How do you connect via IPv6 ?
> > It seems like large packets tunneling issue.
> > Are you able to scp large file via IPv6 ?
> 
> I am using a 6to4 HE tunnel on the client side and native IPv6 on the server side. (It's hosted at Linode that recently enabled native IPv6) 
> 
> You are right. I created just a test.txt file with just hello world in it and i can browse it just fine. I wonder what google is doing because i can browse ipv6.google.com just fine and other test sights. 

They might decrease MTU on their side to allow pass large response from them
to clients.

> Just modified my local MTU to 1280 and it seems to be working now. Sorry for the noise. 


-- 
Igor Sysoev


From nginx-forum at nginx.us  Mon May 16 00:45:00 2011
From: nginx-forum at nginx.us (sprasad90)
Date: Sun, 15 May 2011 16:45:00 -0400
Subject: async n/w logging
In-Reply-To: <4D84CEC7.3030701@chgk.info>
References: <4D84CEC7.3030701@chgk.info>
Message-ID: <0486dd1a2189b48916d561e9f09c58a7.NginxMailingListEnglish@forum.nginx.org>

Hello VT,

i am using ubutu 10.4 Lucid and running nginx 0.8.54. I tried patching
nginx but it fails. see below. I am trying to use syslog-ng to
centralize all the logs to one server. I have had no luck so far.  I
have tried the patches mentioned here
http://wiki.nginx.org/Nginx3rdPartyModules  but no luck so far, they all
error out.  would you have any idea, or point o to a right direction. 
Thank you!

root at nginx-syslog-test:~/nginx-0.8.54# patch -p1 < syslog.patch 
patching file auto/make
patching file auto/options
Hunk #1 succeeded at 118 (offset 8 lines).
Hunk #2 succeeded at 272 (offset 18 lines).
Hunk #3 succeeded at 414 (offset 29 lines).
Hunk #4 succeeded at 433 (offset 32 lines).
patching file auto/summary
Hunk #1 succeeded at 73 with fuzz 1 (offset 2 lines).
patching file src/core/nginx.c
Hunk #2 succeeded at 283 (offset 9 lines).
Hunk #3 succeeded at 409 (offset 19 lines).
patching file src/core/ngx_conf_file.c
Hunk #1 succeeded at 907 with fuzz 1.
patching file src/core/ngx_log.c
Hunk #4 FAILED at 447.
1 out of 4 hunks FAILED -- saving rejects to file
src/core/ngx_log.c.rej
patching file src/http/modules/ngx_http_log_module.c
Hunk #4 succeeded at 835 (offset 18 lines).
Hunk #5 succeeded at 904 (offset 24 lines).

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,183884,198306#msg-198306



From nginx-forum at nginx.us  Mon May 16 00:53:33 2011
From: nginx-forum at nginx.us (sprasad90)
Date: Sun, 15 May 2011 16:53:33 -0400
Subject: syslog patch: please integrate it in the official release/source
In-Reply-To: <99349abca0aa2fdaac0a566af40fe684.NginxMailingListEnglish@forum.nginx.org>
References: <AANLkTin5wd6F7h_Q+dSDN+oAwuXGVtdV=Ng-nyh10AwX@mail.gmail.com>
	<99349abca0aa2fdaac0a566af40fe684.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <58f505d49c41eaebf31e43392393607a.NginxMailingListEnglish@forum.nginx.org>

Hello Daniel,

i am using ubutu 10.4 Lucid and running nginx 0.8.54. I tried patching
nginx with the patch link you mentioned above but it doesn't work.  I
have tried all the patches mentioned in
http://wiki.nginx.org/Nginx3rdPartyModules but none seem to work.  I am
trying to use syslog-ng to centralize all the nginx logs to one server.


After patching, during make i get this 

objs/src/os/unix/ngx_process.o: In function `ngx_process_get_status':
ngx_process.c:(.text+0xe5d): warning: `sys_errlist' is deprecated; use
`strerror' or `strerror_r' instead
ngx_process.c:(.text+0xe4b): warning: `sys_nerr' is deprecated; use
`strerror' or `strerror_r' instead
make[1]: Leaving directory `/root/nginx-0.8.54'

and then after a nginx start,  the file error.log shows up but
access.log is no where to be seen. The error.log doesn't update at all,
so basically the logging doesn't happen at all. 

Any ideas? Any pointers is appreciated.

Thanks!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,148108,198309#msg-198309



From cliff at develix.com  Mon May 16 05:20:38 2011
From: cliff at develix.com (Cliff Wells)
Date: Sun, 15 May 2011 18:20:38 -0700
Subject: Anything wrong with the wiki server?
In-Reply-To: <BANLkTi=EkZGH6Xgwefc8J8ZZmBeXq8oJZg@mail.gmail.com>
References: <BANLkTi=EkZGH6Xgwefc8J8ZZmBeXq8oJZg@mail.gmail.com>
Message-ID: <1305508838.2371.0.camel@portable-evil>

Sorry, had a kernel panic on that system in the wee hours.  It's back up
now.

Cliff

On Sun, 2011-05-15 at 14:46 +0100, Nuno Magalh?es wrote:
> I can connect to nginx.org but not to wiki.nginx.org, times out.
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx




From msosno01 at gmail.com  Mon May 16 05:48:03 2011
From: msosno01 at gmail.com (Maksim Sosnovskiy)
Date: Sun, 15 May 2011 21:48:03 -0400
Subject: no error logging
Message-ID: <BANLkTi=rge06wEHzg9EbwrjuedZn+ca7bA@mail.gmail.com>

I've added an error_log direction to the nginx.conf file, restarted nginx.
However, it's not logging php errors. What could be wrong?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110515/bc36caf3/attachment.html>

From nginx-forum at nginx.us  Mon May 16 06:21:18 2011
From: nginx-forum at nginx.us (jjjx128)
Date: Sun, 15 May 2011 22:21:18 -0400
Subject: no error logging
In-Reply-To: <BANLkTi=rge06wEHzg9EbwrjuedZn+ca7bA@mail.gmail.com>
References: <BANLkTi=rge06wEHzg9EbwrjuedZn+ca7bA@mail.gmail.com>
Message-ID: <863bdb753147028c60fe8f63c57b175f.NginxMailingListEnglish@forum.nginx.org>

Maksim Sosnovskiy Wrote:
-------------------------------------------------------
> I've added an error_log direction to the
> nginx.conf file, restarted nginx.
> However, it's not logging php errors. What could
> be wrong?

Make sure your backend sends errors over fastcgi correctly. Although I'm
not sure if it is even possible with php.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198357,198362#msg-198362



From nbubingo at gmail.com  Mon May 16 07:28:31 2011
From: nbubingo at gmail.com (=?GB2312?B?0qbOsLHz?=)
Date: Mon, 16 May 2011 11:28:31 +0800
Subject: syslog patch: please integrate it in the official release/source
In-Reply-To: <58f505d49c41eaebf31e43392393607a.NginxMailingListEnglish@forum.nginx.org>
References: <AANLkTin5wd6F7h_Q+dSDN+oAwuXGVtdV=Ng-nyh10AwX@mail.gmail.com>
	<99349abca0aa2fdaac0a566af40fe684.NginxMailingListEnglish@forum.nginx.org>
	<58f505d49c41eaebf31e43392393607a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTimXxr1ox6gNdWBQjsBV9Ki1h-r8+Q@mail.gmail.com>

Try my syslog patch(https://github.com/yaoweibin/nginx_syslog_patch), I have
tested it in Nginx-0.8.54 and Nginx-1.0.0. This patch is used in our
production servers.

2011/5/16 sprasad90 <nginx-forum at nginx.us>

> Hello Daniel,
>
> i am using ubutu 10.4 Lucid and running nginx 0.8.54. I tried patching
> nginx with the patch link you mentioned above but it doesn't work.  I
> have tried all the patches mentioned in
> http://wiki.nginx.org/Nginx3rdPartyModules but none seem to work.  I am
> trying to use syslog-ng to centralize all the nginx logs to one server.
>
>
> After patching, during make i get this
>
> objs/src/os/unix/ngx_process.o: In function `ngx_process_get_status':
> ngx_process.c:(.text+0xe5d): warning: `sys_errlist' is deprecated; use
> `strerror' or `strerror_r' instead
> ngx_process.c:(.text+0xe4b): warning: `sys_nerr' is deprecated; use
> `strerror' or `strerror_r' instead
> make[1]: Leaving directory `/root/nginx-0.8.54'
>
> and then after a nginx start,  the file error.log shows up but
> access.log is no where to be seen. The error.log doesn't update at all,
> so basically the logging doesn't happen at all.
>
> Any ideas? Any pointers is appreciated.
>
> Thanks!
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,148108,198309#msg-198309
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110516/4b4bf291/attachment.html>

From mp3geek at gmail.com  Mon May 16 08:01:34 2011
From: mp3geek at gmail.com (Ryan B)
Date: Mon, 16 May 2011 16:01:34 +1200
Subject: Compiling with a custom SSL library
Message-ID: <BANLkTi=zOL_DMRUfvVj97RLvLNMnrZg=pA@mail.gmail.com>

Decided to compile my own copy of openssl (dev snapshot) which is
stored in /usr/local/ssl ,...but nginx isn't detecting it :(

root at server:/usr/local/ssl# ls -al
total 48
drwxr-xr-x  9 root root  4096 2011-05-15 22:15 .
drwxr-xr-x 13 root root  4096 2011-05-15 22:13 ..
drwxr-xr-x  2 root root  4096 2011-05-15 22:15 bin
drwxr-xr-x  2 root root  4096 2011-05-15 22:15 certs
drwxr-xr-x  3 root root  4096 2011-05-15 22:15 include
drwxr-xr-x  4 root root  4096 2011-05-15 22:15 lib
drwxr-xr-x  6 root root  4096 2011-05-15 22:13 man
drwxr-xr-x  2 root root  4096 2011-05-15 22:15 misc
-rw-r--r--  1 root root 10819 2011-05-15 22:15 openssl.cnf
drwxr-xr-x  2 root root  4096 2011-05-15 22:15 private

Tried combinations of --with-openssl=/usr/local/ssl and /usr/local

root at server:~/nginx-1.0.2# make
make -f objs/Makefile
make[1]: Entering directory `/root/nginx-1.0.2'
cd /usr/local/ssl \
        && make clean \
        && ./config --prefix=/usr/local/ssl/.openssl no-shared  no-threads \
        && make \
        && make install LIBDIR=lib
make[2]: Entering directory `/usr/local/ssl'
make[2]: *** No rule to make target `clean'.  Stop.
make[2]: Leaving directory `/usr/local/ssl'
make[1]: *** [/usr/local/ssl/.openssl/include/openssl/ssl.h] Error 2
make[1]: Leaving directory `/root/nginx-1.0.2'
make: *** [build] Error 2

What is the correct ssl dir I need to point it to?


From liulantao at gmail.com  Mon May 16 08:40:40 2011
From: liulantao at gmail.com (Liu Lantao)
Date: Mon, 16 May 2011 12:40:40 +0800
Subject: empty reply from server
In-Reply-To: <8977fb6f50245565eaa9b33541fa9413.NginxMailingListEnglish@forum.nginx.org>
References: <4cd88446bdeea09b413fae442fa115e7.NginxMailingListEnglish@forum.nginx.org>
	<8977fb6f50245565eaa9b33541fa9413.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4DD0AAC8.10906@gmail.com>

how about this file?

                 include /opt/nginx/conf/location_php.conf;



On 05/10/2011 01:18 AM, yellowbox9 wrote:
> Nothing is in the vhosts dir.... and I'm not doing anything like :
>
> error_page 404 =200 /error;
>
> location /error {
> return 444;
> }
>
> Not
>
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,196075,196789#msg-196789
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx


From mp3geek at gmail.com  Mon May 16 08:43:14 2011
From: mp3geek at gmail.com (Ryan B)
Date: Mon, 16 May 2011 16:43:14 +1200
Subject: Compiling with a custom SSL library
In-Reply-To: <BANLkTi=zOL_DMRUfvVj97RLvLNMnrZg=pA@mail.gmail.com>
References: <BANLkTi=zOL_DMRUfvVj97RLvLNMnrZg=pA@mail.gmail.com>
Message-ID: <BANLkTinquO1e62o-K+XnQf+0t8Qx5TGf+w@mail.gmail.com>

Okay, managed to "fix" it, by modifying auto/lib/openssl/conf

from

            CORE_INCS="$CORE_INCS $OPENSSL/.openssl/include"
            CORE_DEPS="$CORE_DEPS $OPENSSL/.openssl/include/openssl/ssl.h"
            CORE_LIBS="$CORE_LIBS $OPENSSL/.openssl/lib/libssl.a"
            CORE_LIBS="$CORE_LIBS $OPENSSL/.openssl/lib/libcrypto.a"

to,

            CORE_INCS="$CORE_INCS $OPENSSL/openssl/include"
            CORE_DEPS="$CORE_DEPS $OPENSSL/openssl/include/openssl/ssl.h"
            CORE_LIBS="$CORE_LIBS $OPENSSL/openssl/lib/libssl.a"
            CORE_LIBS="$CORE_LIBS $OPENSSL/openssl/lib/libcrypto.a"

and link /usr/local/openssl -> /usr/local/ssl

On Mon, May 16, 2011 at 4:01 PM, Ryan B <mp3geek at gmail.com> wrote:
> Decided to compile my own copy of openssl (dev snapshot) which is
> stored in /usr/local/ssl ,...but nginx isn't detecting it :(
>
> root at server:/usr/local/ssl# ls -al
> total 48
> drwxr-xr-x ?9 root root ?4096 2011-05-15 22:15 .
> drwxr-xr-x 13 root root ?4096 2011-05-15 22:13 ..
> drwxr-xr-x ?2 root root ?4096 2011-05-15 22:15 bin
> drwxr-xr-x ?2 root root ?4096 2011-05-15 22:15 certs
> drwxr-xr-x ?3 root root ?4096 2011-05-15 22:15 include
> drwxr-xr-x ?4 root root ?4096 2011-05-15 22:15 lib
> drwxr-xr-x ?6 root root ?4096 2011-05-15 22:13 man
> drwxr-xr-x ?2 root root ?4096 2011-05-15 22:15 misc
> -rw-r--r-- ?1 root root 10819 2011-05-15 22:15 openssl.cnf
> drwxr-xr-x ?2 root root ?4096 2011-05-15 22:15 private
>
> Tried combinations of --with-openssl=/usr/local/ssl and /usr/local
>
> root at server:~/nginx-1.0.2# make
> make -f objs/Makefile
> make[1]: Entering directory `/root/nginx-1.0.2'
> cd /usr/local/ssl \
> ? ? ? ?&& make clean \
> ? ? ? ?&& ./config --prefix=/usr/local/ssl/.openssl no-shared ?no-threads \
> ? ? ? ?&& make \
> ? ? ? ?&& make install LIBDIR=lib
> make[2]: Entering directory `/usr/local/ssl'
> make[2]: *** No rule to make target `clean'. ?Stop.
> make[2]: Leaving directory `/usr/local/ssl'
> make[1]: *** [/usr/local/ssl/.openssl/include/openssl/ssl.h] Error 2
> make[1]: Leaving directory `/root/nginx-1.0.2'
> make: *** [build] Error 2
>
> What is the correct ssl dir I need to point it to?
>


From igor at sysoev.ru  Mon May 16 09:06:49 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Mon, 16 May 2011 09:06:49 +0400
Subject: Compiling with a custom SSL library
In-Reply-To: <BANLkTi=zOL_DMRUfvVj97RLvLNMnrZg=pA@mail.gmail.com>
References: <BANLkTi=zOL_DMRUfvVj97RLvLNMnrZg=pA@mail.gmail.com>
Message-ID: <20110516050649.GA24536@sysoev.ru>

On Mon, May 16, 2011 at 04:01:34PM +1200, Ryan B wrote:
> Decided to compile my own copy of openssl (dev snapshot) which is
> stored in /usr/local/ssl ,...but nginx isn't detecting it :(
> 
> root at server:/usr/local/ssl# ls -al
> total 48
> drwxr-xr-x  9 root root  4096 2011-05-15 22:15 .
> drwxr-xr-x 13 root root  4096 2011-05-15 22:13 ..
> drwxr-xr-x  2 root root  4096 2011-05-15 22:15 bin
> drwxr-xr-x  2 root root  4096 2011-05-15 22:15 certs
> drwxr-xr-x  3 root root  4096 2011-05-15 22:15 include
> drwxr-xr-x  4 root root  4096 2011-05-15 22:15 lib
> drwxr-xr-x  6 root root  4096 2011-05-15 22:13 man
> drwxr-xr-x  2 root root  4096 2011-05-15 22:15 misc
> -rw-r--r--  1 root root 10819 2011-05-15 22:15 openssl.cnf
> drwxr-xr-x  2 root root  4096 2011-05-15 22:15 private
> 
> Tried combinations of --with-openssl=/usr/local/ssl and /usr/local
> 
> root at server:~/nginx-1.0.2# make
> make -f objs/Makefile
> make[1]: Entering directory `/root/nginx-1.0.2'
> cd /usr/local/ssl \
>         && make clean \
>         && ./config --prefix=/usr/local/ssl/.openssl no-shared  no-threads \
>         && make \
>         && make install LIBDIR=lib
> make[2]: Entering directory `/usr/local/ssl'
> make[2]: *** No rule to make target `clean'.  Stop.
> make[2]: Leaving directory `/usr/local/ssl'
> make[1]: *** [/usr/local/ssl/.openssl/include/openssl/ssl.h] Error 2
> make[1]: Leaving directory `/root/nginx-1.0.2'
> make: *** [build] Error 2
> 
> What is the correct ssl dir I need to point it to?

./configure --help | grep openssl=
  --with-openssl=DIR                 set path to OpenSSL library sources

The sources, but not pre-built installation.

wget http://openssl.org/source/openssl-1.0.0d.tar.gz
tar zxf openssl-1.0.0d.tar.gz
cd nginx-1.0.2
./configure --with-openssl=../openssl-1.0.0d ...


-- 
Igor Sysoev


From nunomagalhaes at eu.ipp.pt  Mon May 16 12:21:32 2011
From: nunomagalhaes at eu.ipp.pt (=?UTF-8?Q?Nuno_Magalh=C3=A3es?=)
Date: Mon, 16 May 2011 09:21:32 +0100
Subject: Gzip vs Gzip static
Message-ID: <BANLkTikXdNiAvgUyQDzc_8eVU6hhZsWBFQ@mail.gmail.com>

Hi,

This may be a silly question, but, does gzip static depend/extend
gzip, or can i just compile with gzip static and without gzip? I would
assume either gzip or gzip+static, but i'd like to confirm.

Thanks,
Nuno


From igor at sysoev.ru  Mon May 16 12:25:19 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Mon, 16 May 2011 12:25:19 +0400
Subject: Gzip vs Gzip static
In-Reply-To: <BANLkTikXdNiAvgUyQDzc_8eVU6hhZsWBFQ@mail.gmail.com>
References: <BANLkTikXdNiAvgUyQDzc_8eVU6hhZsWBFQ@mail.gmail.com>
Message-ID: <20110516082519.GC33895@sysoev.ru>

On Mon, May 16, 2011 at 09:21:32AM +0100, Nuno Magalh?es wrote:
> Hi,
> 
> This may be a silly question, but, does gzip static depend/extend
> gzip, or can i just compile with gzip static and without gzip? I would
> assume either gzip or gzip+static, but i'd like to confirm.

gzip_static does not depend on gzip filter. You can use gzip_static
without compiled gzip filter.


-- 
Igor Sysoev


From Pekka.Panula at sofor.fi  Mon May 16 14:14:27 2011
From: Pekka.Panula at sofor.fi (Pekka.Panula at sofor.fi)
Date: Mon, 16 May 2011 13:14:27 +0300
Subject: Got timeout problems, even with 1.0.2, without old patches,
	reverse proxy with cache features
Message-ID: <OF4A098C99.0C33F168-ONC2257892.0037D2BD-C2257892.003841C1@sofor.fi>

Hi

Even with 1.0.2 version i am needed to patch these 2 patches: 
http://nginx.org/pipermail/nginx-devel/2010-January/000100.html
Because i am getting timeouts without those 2 patch. My backend is IBM 
Lotus Domino and these are cases where body is 0 bytes but there is 
redirection / location on headers, it seems nginx on default stalled (when 
using cache features), it does not close connection between client and 
nginx.

Seems to be fixed if i patch nginx with those patches, perhaps integrate 
them to nginx permanently?

Terveisin/Regards,
   Pekka Panula, Sofor Oy - Jatkuvat palvelut
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110516/5854737e/attachment.html>

From mdounin at mdounin.ru  Mon May 16 14:29:24 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 16 May 2011 14:29:24 +0400
Subject: Got timeout problems, even with 1.0.2, without old patches,
	reverse proxy with cache features
In-Reply-To: <OF4A098C99.0C33F168-ONC2257892.0037D2BD-C2257892.003841C1@sofor.fi>
References: <OF4A098C99.0C33F168-ONC2257892.0037D2BD-C2257892.003841C1@sofor.fi>
Message-ID: <20110516102924.GE42265@mdounin.ru>

Hello!

On Mon, May 16, 2011 at 01:14:27PM +0300, Pekka.Panula at sofor.fi wrote:

> Hi
> 
> Even with 1.0.2 version i am needed to patch these 2 patches: 
> http://nginx.org/pipermail/nginx-devel/2010-January/000100.html
> Because i am getting timeouts without those 2 patch. My backend is IBM 
> Lotus Domino and these are cases where body is 0 bytes but there is 
> redirection / location on headers, it seems nginx on default stalled (when 
> using cache features), it does not close connection between client and 
> nginx.
> 
> Seems to be fixed if i patch nginx with those patches, perhaps integrate 
> them to nginx permanently?

JFYI, more recent versions of these patches may be found here:

http://nginx.org/pipermail/nginx-devel/2011-February/000743.html
http://nginx.org/pipermail/nginx-devel/2011-February/000744.html

Maxim Dounin


From mdounin at mdounin.ru  Mon May 16 15:08:34 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 16 May 2011 15:08:34 +0400
Subject: gzip compression not taken into account?
In-Reply-To: <e1eb6064bdd055b2c4aca6d4471692ca.NginxMailingListEnglish@forum.nginx.org>
References: <a1bb7953bf979c5dba6b2802f7e5b67a.NginxMailingListEnglish@forum.nginx.org>
	<0c89e375c17616d0cdc09c2d633194dd.NginxMailingListEnglish@forum.nginx.org>
	<e1eb6064bdd055b2c4aca6d4471692ca.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110516110834.GG42265@mdounin.ru>

Hello!

On Sat, May 14, 2011 at 05:51:33PM -0400, jfix wrote:

> Thanks Thomas for your reply.
> 
> I've redone the test by explicitly supplying the proxy, and indeed it
> seems to remove the Accept-encoding header, as even with --compressed
> option, the response is now uncompressed:
> 
> M:>curl --compressed -v -x wsg-proxy.example.org:80
> http://example.org/resource/style/global.css
> * About to connect() to proxy wsg-proxy.example.org port 80 (#0)
> *   Trying 10.101.128.15... connected
> * Connected to wsg-proxy.oecd.org (10.101.128.15) port 80 (#0)
> > GET http://example.org/resource/style/global.css HTTP/1.1
> > User-Agent: curl/7.19.7 (i386-pc-win32) libcurl/7.19.7 OpenSSL/0.9.8l
> zlib/1.2.3
> > Host: example.org
> > Accept: */*
> > Accept-Encoding: deflate, gzip
> > Proxy-Connection: Keep-Alive
> >
> < HTTP/1.1 200 OK
> < Server: nginx
> < Date: Sat, 14 May 2011 21:46:49 GMT
> < Content-Type: text/css
> < Vary: Accept-Encoding
> < Last-Modified: Fri, 13 May 2011 16:00:36 GMT
> < Expires: Thu, 31 Dec 2037 23:55:55 GMT
> < Cache-Control: max-age=315360000
> < Content-Length: 43345
> < Age: 0
> < Via: 1.1 localhost.localdomain
> 
> So, it looks like I have to talk to my company's IT people ... Thanks
> for your help.

Please note that by default nginx only returns compressed response 
to HTTP/1.1 requests.  By using proxy you may be in fact using 
HTTP/1.0 in request to nginx.

Additionally, nginx will look though request headers to find out 
if request was from proxy, and by default won't return compressed 
response to proxies.  With gzip_proxied as set in your config it 
won't return response in question compressed as well.

See here for more details:

http://wiki.nginx.org/HttpGzipModule#gzip_http_version
http://wiki.nginx.org/HttpGzipModule#gzip_proxied

Please also note that unconditionally enabling gzip for everything 
may not be a good idea.  If enabling gzip for proxied requests, 
please at least make sure that (a) gzip_vary set to on and (b) 
gzip for HTTP/1.0 is disabled (default).

Maxim Dounin


From nginx-forum at nginx.us  Mon May 16 15:29:33 2011
From: nginx-forum at nginx.us (jfix)
Date: Mon, 16 May 2011 07:29:33 -0400
Subject: gzip compression not taken into account?
In-Reply-To: <20110516110834.GG42265@mdounin.ru>
References: <20110516110834.GG42265@mdounin.ru>
Message-ID: <6ef2bc692a4a9c4f02e0c9a85855623a.NginxMailingListEnglish@forum.nginx.org>

Hi Maxim, thanks for your input.

I had, in the meantime, added the directives

gzip_http_version 1.0
gzip_vary on


Does the fact that the response contains this header

Via:1.1 localhost.localdomain

mean that HTTP 1.1 is being used?  Having changed temporarily the
gzip_proxied directive to

gzip_proxied any

hasn't changed a thing, unfortunately.  I will look in more detail at
the gzip_proxied values and see how they are applicable or not.

Thanks,
Jakob.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,197898,198433#msg-198433



From mdounin at mdounin.ru  Mon May 16 15:33:12 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 16 May 2011 15:33:12 +0400
Subject: Keepalived Connections Reset after reloading the configuration
	(HUP Signal)
In-Reply-To: <46781f68e473c0d6a12e4a3e9788a38e.NginxMailingListEnglish@forum.nginx.org>
References: <46781f68e473c0d6a12e4a3e9788a38e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110516113312.GH42265@mdounin.ru>

Hello!

On Fri, May 13, 2011 at 07:24:12PM -0400, Jocelyn Mocquant wrote:

> Hello,
> 
>    I'm dealing with a problem. When reloading the nginx configuration,
> all keepalived connections receive the  TCP reset flag after I send a
> HUP signal to the master process. If I comment the line responsible for
> enabling the keepalive feature in the configuration, the problem
> disappear (nginx version is 0.9.7). 

On configuration reloading nginx starts new worker processes with 
new configuration and old worker processes are terminated as soon 
as they finish processing of requests.  All keepalive connections 
in old workers are closed accordingly.  HTTP/1.1 clients are 
required to handle keepalive connection close, so this shoudln't 
be a problem.

It's not clear why you see RST instead of normal close with 
FIN/FIN+ACK/ACK (are you?), but this shouldn't the problem by 
itself anyway.

Maxim Dounin


From tomlove at gmail.com  Mon May 16 17:57:43 2011
From: tomlove at gmail.com (Thomas Love)
Date: Mon, 16 May 2011 15:57:43 +0200
Subject: gzip compression not taken into account?
In-Reply-To: <6ef2bc692a4a9c4f02e0c9a85855623a.NginxMailingListEnglish@forum.nginx.org>
References: <20110516110834.GG42265@mdounin.ru>
	<6ef2bc692a4a9c4f02e0c9a85855623a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTinGejVYJacdo_R0hj56HNf1edDdAA@mail.gmail.com>

That you've added those directives with no fix makes me more confident that
your proxy is simply not forwarding your Accept-Encoding: gzip header.

I'm not sure if nginx can log request headers, but if you have it configured
for PHP, you can request a simple test script:

var_dump($_SERVER["HTTP_ACCEPT_ENCODING"]);


Thomas




On 16 May 2011 13:29, jfix <nginx-forum at nginx.us> wrote:

> Hi Maxim, thanks for your input.
>
> I had, in the meantime, added the directives
>
> gzip_http_version 1.0
> gzip_vary on
>
>
> Does the fact that the response contains this header
>
> Via:1.1 localhost.localdomain
>
> mean that HTTP 1.1 is being used?  Having changed temporarily the
> gzip_proxied directive to
>
> gzip_proxied any
>
> hasn't changed a thing, unfortunately.  I will look in more detail at
> the gzip_proxied values and see how they are applicable or not.
>
> Thanks,
> Jakob.
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,197898,198433#msg-198433
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110516/6adf7ed6/attachment.html>

From mdounin at mdounin.ru  Mon May 16 18:25:35 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 16 May 2011 18:25:35 +0400
Subject: nginx 1.0.2 intermittent segfault error 4
In-Reply-To: <041aa0fb00af03026d71bcb288c84db7.NginxMailingListEnglish@forum.nginx.org>
References: <041aa0fb00af03026d71bcb288c84db7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110516142535.GK42265@mdounin.ru>

Hello!

On Sat, May 14, 2011 at 11:16:22AM -0400, inph wrote:

> Hi,
> 
> I've noticed some nginx segfaults over the last week or so with 1.0.1
> and 1.0.2. I havent tested any other version for long enough.
> The segfaults are not immediate and sometimes it requires several hours
> to reproduce through normal use.
> 
> Here is the debug info, if you require anymore information please ask.
> 
> [12702.705897] nginx[1769]: segfault at 50545448 ip 0807d1fa sp bfc241d0
> error 4 in nginx[8048000+82000]

[...]

>    location ~ \.php$ {
>         try_files      $uri =404;
>         access_log     /var/log/nginx/upstream.log  upstream;
>         fastcgi_pass   unix:/var/run/php-fpm.sock;
>         fastcgi_index  index.php;
>         include        fastcgi.conf;
>         fastcgi_intercept_errors        on;
>         fastcgi_ignore_client_abort     on;

Yes, thank you for your report.  The problem is with 
fastcgi_ignore_client_abort, it doesn't handle POST requests 
correctly.

Switching fastcgi_ignore_client_abort back to default (off) should 
resolve this issue.

Maxim Dounin


From jocelyn.mocquant at gmail.com  Tue May 17 02:12:05 2011
From: jocelyn.mocquant at gmail.com (Jocelyn)
Date: Mon, 16 May 2011 15:12:05 -0700
Subject: Keepalived Connections Reset after reloading the configuration
	(HUP Signal)
In-Reply-To: <20110516113312.GH42265@mdounin.ru>
References: <46781f68e473c0d6a12e4a3e9788a38e.NginxMailingListEnglish@forum.nginx.org>
	<20110516113312.GH42265@mdounin.ru>
Message-ID: <BANLkTi=wW3F1S4V7tq1-TrWyutW9XatPGQ@mail.gmail.com>

Thanks Maxim for your answer,

If I remove the keep alive directive, all happens as expected. Apparently,
sending USR2 and WINCH solve the problem.

--
Regards
       Jocelyn


On Mon, May 16, 2011 at 4:33 AM, Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
>
> On Fri, May 13, 2011 at 07:24:12PM -0400, Jocelyn Mocquant wrote:
>
> > Hello,
> >
> >    I'm dealing with a problem. When reloading the nginx configuration,
> > all keepalived connections receive the  TCP reset flag after I send a
> > HUP signal to the master process. If I comment the line responsible for
> > enabling the keepalive feature in the configuration, the problem
> > disappear (nginx version is 0.9.7).
>
> On configuration reloading nginx starts new worker processes with
> new configuration and old worker processes are terminated as soon
> as they finish processing of requests.  All keepalive connections
> in old workers are closed accordingly.  HTTP/1.1 clients are
> required to handle keepalive connection close, so this shoudln't
> be a problem.
>
> It's not clear why you see RST instead of normal close with
> FIN/FIN+ACK/ACK (are you?), but this shouldn't the problem by
> itself anyway.
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110516/986c3034/attachment-0001.html>

From nginx-forum at nginx.us  Tue May 17 04:01:55 2011
From: nginx-forum at nginx.us (George)
Date: Mon, 16 May 2011 20:01:55 -0400
Subject: Compiling with a custom SSL library
In-Reply-To: <20110516050649.GA24536@sysoev.ru>
References: <20110516050649.GA24536@sysoev.ru>
Message-ID: <8ae970b3654f137ab8bb7cd163c614f9.NginxMailingListEnglish@forum.nginx.org>

thanks Igor that helped me with my custom nginx 1.0.2 + php-fpm openssl
1.0.0d install as well :)

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198369,198600#msg-198600



From nginx-forum at nginx.us  Tue May 17 10:58:53 2011
From: nginx-forum at nginx.us (Mike_1)
Date: Tue, 17 May 2011 02:58:53 -0400
Subject: smtp proxy to external smtp server
Message-ID: <b8a7e4e0d4c5e07086ae3e1db4d2ad45.NginxMailingListEnglish@forum.nginx.org>

Hi all

Can nginx used as smtp proxy connecting to an external smtp server?
The external smtp server requires smtp auth.
--
Thanks
Mike

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198667,198667#msg-198667



From lists at ruby-forum.com  Tue May 17 18:31:45 2011
From: lists at ruby-forum.com (Antoine L.)
Date: Tue, 17 May 2011 16:31:45 +0200
Subject: 504 error Gateway Timeout.
In-Reply-To: <AANLkTikH+rAOaFuAPA7q0SzcHns9RRn099oqrFevvCtK@mail.gmail.com>
References: <AANLkTikH+rAOaFuAPA7q0SzcHns9RRn099oqrFevvCtK@mail.gmail.com>
Message-ID: <4724c2b955f220f6f9cfde53e5ca0e5e@ruby-forum.com>

Hi guys,

I have the exactly same issue.
What about include proxy.inc ?

Those are my values :
proxy_connect_timeout 90;
proxy_send_timeout 90;
proxy_read_timeout 90;
proxy_buffer_size 4k;
proxy_buffers 4 32k;
proxy_busy_buffers_size 64k;
proxy_temp_file_write_size 64k;

Use it with Magento on a VPS and always need to restart APACHE

Thanks for the help,

Antoine

-- 
Posted via http://www.ruby-forum.com/.


From nginx-forum at nginx.us  Tue May 17 20:33:14 2011
From: nginx-forum at nginx.us (arleybls)
Date: Tue, 17 May 2011 12:33:14 -0400
Subject: Profiling NGINX
Message-ID: <7af7092b35bc728bf38cf1582c0e452e.NginxMailingListEnglish@forum.nginx.org>

Hello all,

I need to profile nginx+own_module. I've being trying to use gprof, but
even tough compiling along with --with-cc-opt="-pg" --with-ld-opt="-pg"
the gmon.out is not being generated. 

Any ideas?

Cheers,
AB

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198795,198795#msg-198795



From nginx-forum at nginx.us  Tue May 17 21:25:11 2011
From: nginx-forum at nginx.us (arleybls)
Date: Tue, 17 May 2011 13:25:11 -0400
Subject: Profiling NGINX
In-Reply-To: <7af7092b35bc728bf38cf1582c0e452e.NginxMailingListEnglish@forum.nginx.org>
References: <7af7092b35bc728bf38cf1582c0e452e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <ec4c849a390219512e5962ce49a648cf.NginxMailingListEnglish@forum.nginx.org>

arleybls Wrote:
-------------------------------------------------------
> Hello all,
> 
> I need to profile nginx+own_module. I've being
> trying to use gprof, but even tough compiling
> along with --with-cc-opt="-pg" --with-ld-opt="-pg"
> the gmon.out is not being generated. 
> 
> Any ideas?
> 
> Cheers,
> AB

Nerve mind, running manually along with "daemon  off;" and
"master_process  off;" did the trick.

Ty,
AB

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198795,198812#msg-198812



From nginx-forum at nginx.us  Tue May 17 22:05:41 2011
From: nginx-forum at nginx.us (lkmcmillen)
Date: Tue, 17 May 2011 14:05:41 -0400
Subject: CanNginx Mail Server replace Postfix to listen only?
Message-ID: <8f18130d95dc425c4b1bb8fb940f0c85.NginxMailingListEnglish@forum.nginx.org>

Can the Nginx mail server be used in replacement of Postfix?

I am a bit new to this.  However, I have a small, fully functioning,
unmanaged VPS server, rented from Linode.com, which is setup as a LEMP
server, running Ubuntu 11.04, Nginx 1.0.2, MariaDB 5.2.6, PHP 5.3.6
(PHP-FPM), xcache 1.3.1, vBulletin 4.1.2, and Postfix which was setup
using the following script:

###########################################################
# Postfix
###########################################################

	# Installs postfix and configure to listen only on the local interface.
Also
	# allows for local mail delivery

	{
	echo "postfix postfix/main_mailer_type select Internet Site" |
debconf-set-selections
	echo "postfix postfix/mailname string localhost" |
debconf-set-selections
	echo "postfix postfix/destinations string localhost.localdomain,
localhost" | debconf-set-selections
	aptitude -y install postfix
	/usr/sbin/postconf -e "inet_interfaces = loopback-only"
	#/usr/sbin/postconf -e "local_transport = error:local delivery is
disabled"

	touch /tmp/restart-postfix
	}

############################################################

Can the Nginx mail server be used in the same manner in replacement of
Postfix?

If so:

(1) Is there good reason not to use the Nginx mail server in such a
manner?

(2) What do I need to do in order to use the Nginx mail server instead
of Postfix?

As a matter of information, I recently subscribed to Google Apps for the
hosted domain. We need Postfix, or the Nginx mail server, only to send
and receive email from our web applications.

Thank you.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198821,198821#msg-198821



From igor at sysoev.ru  Tue May 17 22:39:12 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 17 May 2011 22:39:12 +0400
Subject: CanNginx Mail Server replace Postfix to listen only?
In-Reply-To: <8f18130d95dc425c4b1bb8fb940f0c85.NginxMailingListEnglish@forum.nginx.org>
References: <8f18130d95dc425c4b1bb8fb940f0c85.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110517183912.GA18234@sysoev.ru>

On Tue, May 17, 2011 at 02:05:41PM -0400, lkmcmillen wrote:
> Can the Nginx mail server be used in replacement of Postfix?

No.


-- 
Igor Sysoev


From mitchell.hashimoto at gmail.com  Wed May 18 00:14:30 2011
From: mitchell.hashimoto at gmail.com (Mitchell Hashimoto)
Date: Tue, 17 May 2011 13:14:30 -0700
Subject: Conditionally set header in proxy
Message-ID: <BANLkTikvQOSbvfsWTtOHNLjWj8+N37jhDQ@mail.gmail.com>

Hello,

I'm doing SSL termination in nginx, and I'd like to set a header such as
"X-Sheme https" ONLY if X-Scheme is not set. The reason is because an
upstream load balancer may have terminated SSL as well, and I don't want to
overwrite that value.

Is this possible?

Best,
Mitchell
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110517/59b97314/attachment.html>

From nginx-forum at nginx.us  Wed May 18 02:51:33 2011
From: nginx-forum at nginx.us (the_traveller)
Date: Tue, 17 May 2011 18:51:33 -0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <2507a2f6fd2c5e8a743f33ca869dfa82.NginxMailingListEnglish@forum.nginx.org>
References: <72a81ff3e80ed0b4b78cb81a348826cf@ruby-forum.com>
	<2507a2f6fd2c5e8a743f33ca869dfa82.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <f97ed64e3fc4a8c4611fc3a81c18c3b7.NginxMailingListEnglish@forum.nginx.org>

First of all thank you very much mr Igor This is a great help for me
but as i'm new to nginx  so i'm little bit careful
but i have afew more things to ask about

do i have to make any defination to these params

SCRIPT_FILENAME
maindir

which is used as
fastcgi_param SCRIPT_FILENAME /path/to/script/pages.php;
AND
location /maindir/chat/

Or they just have to be written as they are??

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,23868,198889#msg-198889



From nginx-forum at nginx.us  Wed May 18 08:09:34 2011
From: nginx-forum at nginx.us (arti)
Date: Wed, 18 May 2011 00:09:34 -0400
Subject: File AIO and multiple nginx instances
Message-ID: <deaa25c9a3b732a5c67d62a2a7338373.NginxMailingListEnglish@forum.nginx.org>

Hello.

I need to run two nginx instances on one server (not two workers, but
two master processes with their own number of workers each). The problem
is that when nginx is compiled with aio support I can start only one
instance. When I try to start the second one, only master process starts
and these lines appear in error log:

[emerg] 12789#0: io_setup() failed (1: Operation not permitted)
...
[alert] 12715#0: worker process 12789 exited with fatal code 2 and can
not be respawn
...

This happens even if aio directive isn't used in configuration.

My system is CentOS 5.5 with kernel 2.6.18-194.32.1.el5, I've tried
nginx 1.0.0 and nginx 1.0.2

So, is it expected behavior (I don't know nginx and/or aio internals),
or is it a bug in nginx or aio, and is there a way to start two nginx
instances, both with aio support?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198945,198945#msg-198945



From igor at sysoev.ru  Wed May 18 08:43:36 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 18 May 2011 08:43:36 +0400
Subject: Conditionally set header in proxy
In-Reply-To: <BANLkTikvQOSbvfsWTtOHNLjWj8+N37jhDQ@mail.gmail.com>
References: <BANLkTikvQOSbvfsWTtOHNLjWj8+N37jhDQ@mail.gmail.com>
Message-ID: <20110518044336.GA45611@sysoev.ru>

On Tue, May 17, 2011 at 01:14:30PM -0700, Mitchell Hashimoto wrote:
> Hello,
> 
> I'm doing SSL termination in nginx, and I'd like to set a header such as
> "X-Sheme https" ONLY if X-Scheme is not set. The reason is because an
> upstream load balancer may have terminated SSL as well, and I don't want to
> overwrite that value.
> 
> Is this possible?

nginx-0.9.0+:

http {
    map  $http_x_scheme  $x_scheme {
         ""              https;
         default         $http_x_scheme;
    }

    server {

        location / {

            proxy_set_header  X-Scheme  $x_scheme;


-- 
Igor Sysoev


From nginx-forum at nginx.us  Wed May 18 09:13:39 2011
From: nginx-forum at nginx.us (felka)
Date: Wed, 18 May 2011 01:13:39 -0400
Subject: Proxy module - upstream server recovery from inoperative mode
Message-ID: <b32bb6e771b0dab82ea2a7b935bf7ce1.NginxMailingListEnglish@forum.nginx.org>

Hi All,

We are trying to use NGINX as a load balancing solution for our
production servers. Our infrastructure contains few application server
using SOAP XML over HTTP. NGNIX suppose to balance between servers.
Please see config:

user              nginx;
worker_processes  10;
error_log  /var/log/nginx/error.log;
pid        /var/run/nginx.pid;

events {
    worker_connections  2048;
}

http {
        upstream CS {
                server 70.70.0.31:80 max_fails=1 fail_timeout=1s;
                server 70.70.0.41:80 max_fails=1 fail_timeout=1s;
        }
        server {
                listen 54321;
                location / {
                proxy_pass http://CS;
                proxy_read_timeout 2s;
        #       proxy_connect_timeout 200ms;
                }
        }
}
 
NGINX does the work great until some of backend server fails to answer
to NGINX in time and it send request to other server. Until this point
everything is great and working. Now failed server should recover from
inoperative mode to operetive mode after 1 second as I've defined
fail_timeout to 1 second (Here might be my mistake because it may be
recovers but right after it fails again) and it doesn't. I'm getting to
the point where NGINX says that there is no operative upstream servers
available because in some point load to the servers is so high that both
of servers unable to answer in time. 
We are working on solving our servivce issue. However I think it is
cruicial for load balancer be able moving failed upstream server to
operative mode. May be I'm doing something wrong. Which log level will
show the recovery?


I will appreciate for any help!

Regards,
Seva Feldman

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198952,198952#msg-198952



From igor at sysoev.ru  Wed May 18 10:49:21 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 18 May 2011 10:49:21 +0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <f97ed64e3fc4a8c4611fc3a81c18c3b7.NginxMailingListEnglish@forum.nginx.org>
References: <72a81ff3e80ed0b4b78cb81a348826cf@ruby-forum.com>
	<2507a2f6fd2c5e8a743f33ca869dfa82.NginxMailingListEnglish@forum.nginx.org>
	<f97ed64e3fc4a8c4611fc3a81c18c3b7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110518064921.GC45611@sysoev.ru>

On Tue, May 17, 2011 at 06:51:33PM -0400, the_traveller wrote:
> First of all thank you very much mr Igor This is a great help for me
> but as i'm new to nginx  so i'm little bit careful
> but i have afew more things to ask about
> 
> do i have to make any defination to these params
> 
> SCRIPT_FILENAME
> maindir

"/maindir/" is from
RewriteBase /maindir/

By default it is "/", for example,

location /chat/

SCRIPT_FILENAME is FastCGI parameter which contains file system script name.

> which is used as
> fastcgi_param SCRIPT_FILENAME /path/to/script/pages.php;
> AND
> location /maindir/chat/
> 
> Or they just have to be written as they are??


-- 
Igor Sysoev


From igor at sysoev.ru  Wed May 18 11:34:01 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 18 May 2011 11:34:01 +0400
Subject: File AIO and multiple nginx instances
In-Reply-To: <deaa25c9a3b732a5c67d62a2a7338373.NginxMailingListEnglish@forum.nginx.org>
References: <deaa25c9a3b732a5c67d62a2a7338373.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110518073401.GG45611@sysoev.ru>

On Wed, May 18, 2011 at 12:09:34AM -0400, arti wrote:
> Hello.
> 
> I need to run two nginx instances on one server (not two workers, but
> two master processes with their own number of workers each). The problem
> is that when nginx is compiled with aio support I can start only one
> instance. When I try to start the second one, only master process starts
> and these lines appear in error log:
> 
> [emerg] 12789#0: io_setup() failed (1: Operation not permitted)
> ...
> [alert] 12715#0: worker process 12789 exited with fatal code 2 and can
> not be respawn
> ...
> 
> This happens even if aio directive isn't used in configuration.
> 
> My system is CentOS 5.5 with kernel 2.6.18-194.32.1.el5, I've tried
> nginx 1.0.0 and nginx 1.0.2
> 
> So, is it expected behavior (I don't know nginx and/or aio internals),
> or is it a bug in nginx or aio, and is there a way to start two nginx
> instances, both with aio support?

It's strange, but io_setup(2) man does not even mention EPERM:
http://www.kernel.org/doc/man-pages/online/pages/man2/io_setup.2.html


-- 
Igor Sysoev


From pascal.nitsche at fansubcode.org  Wed May 18 11:49:40 2011
From: pascal.nitsche at fansubcode.org (Pascal Nitsche)
Date: Wed, 18 May 2011 09:49:40 +0200
Subject: ngx_http_secure_download report back to application?
Message-ID: <4DD37A14.2050201@fansubcode.org>

Hello,

I'm thinking about writing an application, which should use nginx' 
secure download module to generate secure - expireing - links and limit 
the number of (successful) downloads. Since it only should count 
successful downloads it needs to know if the download was successful so 
it would be good to have nginx report to the application "Hey, I've got 
one more complete download". In the best case also "Download failed, but 
I transfered N bytes" for statistical reasons inside the application 
(This could be used to track traffic usage by file and stop generating 
links if a traffic limit has passed).

The download server and the application server could - but haven't to - 
run on different machines.

Now my question is: Is something like this possible? Can I send a 
response to an application in - let's say XML or JSON? And if yes: How?

Thanks in advance,

Pascal Nitsche


From wubingzheng at 163.com  Wed May 18 13:15:00 2011
From: wubingzheng at 163.com (=?UTF-8?Q?Wu=C2=A0Bingzheng?=)
Date: Wed, 18 May 2011 17:15:00 +0800 (CST)
Subject: 2 questions on send_timeout (maybe bugs)
Message-ID: <10a6803e.1907c.130026300cb.Coremail.wubingzheng@163.com>

hi all,


I meet 2 question when using nginx 1.0.0.


Question1.

what happens if the downstream connection breaks, before all data sent to the client?

1. get client's request
2. set handler: r->read_event_handle=ngx_http_block_reading
3. send data to client, and blocked ( write_filter.c:c->send_chain() returns non-NULL or non-error), and go back to epoll_wait
4. HERE, the downstream breaks
5. epoll detect the breaking, so epoll_wait() continues in ngx_epoll_process_events(), where the revents==EPOLL_IN | EPOLL_ERR | EPOLL_HUP. So we call rev->handler(), which calls r->read_event_handle, which dose nothing. AS A RESULT, the downstream breaking was not detected by nginx !!!
6. Then, we have to wait for send_timeout...

This situation dosenot occur everytime. If the downstream breaks, when nginx is sending data, the breaking will be detected.
But I can make this happen all the time, when:
1. the client and server are far away from each other (e.g. not in a local net)
2. using memc module



Question2.

I set 'send_timeout' as 20s, but nginx timeouts in 40s actually.

1. in src/http/ngx_http_upstream.c[line2448]:ngx_http_upstream_process_non_buffered_request(), a timer is added;
2. after that timer timeout, in src/http/ngx_http_request.c[line2179]:ngx_http_set_write_handler(), another timer is added again.

so the send_timeout is doubled.




Thanks
Wu


From nginx-forum at nginx.us  Wed May 18 17:03:40 2011
From: nginx-forum at nginx.us (Ensiferous)
Date: Wed, 18 May 2011 09:03:40 -0400
Subject: ngx_http_secure_download report back to application?
In-Reply-To: <4DD37A14.2050201@fansubcode.org>
References: <4DD37A14.2050201@fansubcode.org>
Message-ID: <06d9bc75fdbcf1ed50b63659bdf8cc26.NginxMailingListEnglish@forum.nginx.org>

Hi,

You can do this with http://wiki.nginx.org/HttpCoreModule#post_action
and http://wiki.nginx.org/HttpCoreModule#.24request_completion

Sincerely,
Martin Fjordvald

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198967,199037#msg-199037



From mdounin at mdounin.ru  Wed May 18 17:35:07 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Wed, 18 May 2011 17:35:07 +0400
Subject: 2 questions on send_timeout (maybe bugs)
In-Reply-To: <10a6803e.1907c.130026300cb.Coremail.wubingzheng@163.com>
References: <10a6803e.1907c.130026300cb.Coremail.wubingzheng@163.com>
Message-ID: <20110518133507.GV42265@mdounin.ru>

Hello!

On Wed, May 18, 2011 at 05:15:00PM +0800, Wu Bingzheng wrote:

> hi all,
> 
> 
> I meet 2 question when using nginx 1.0.0.
> 
> 
> Question1.
> 
> what happens if the downstream connection breaks, before all 
> data sent to the client?
> 
> 1. get client's request
> 2. set handler: r->read_event_handle=ngx_http_block_reading
> 3. send data to client, and blocked ( 
> write_filter.c:c->send_chain() returns non-NULL or non-error), 
> and go back to epoll_wait
> 4. HERE, the downstream breaks
> 5. epoll detect the breaking, so epoll_wait() continues in 
> ngx_epoll_process_events(), where the revents==EPOLL_IN | 
> EPOLL_ERR | EPOLL_HUP. So we call rev->handler(), which calls 
> r->read_event_handle, which dose nothing. AS A RESULT, the 
> downstream breaking was not detected by nginx !!!

Here write event handler will be called.  While standard 
ngx_http_writer() is in use - it will call output filter chain and 
detect error.  Else it's up to active handler to correctly handle 
this, e.g. standard upstream modules (proxy/fastcgi/...) may 
either terminate request immediately or wait for full upstream 
response depending on various settings (proxy_ignore_client_abort, 
proxy_cache/proxy_store enabled).

> 6. Then, we have to wait for send_timeout...
> 
> This situation dosenot occur everytime. If the downstream 
> breaks, when nginx is sending data, the breaking will be 
> detected.
> But I can make this happen all the time, when:
> 1. the client and server are far away from each other (e.g. not 
> in a local net)
> 2. using memc module
>
> Question2.
> 
> I set 'send_timeout' as 20s, but nginx timeouts in 40s actually.
> 
> 1. in 
> src/http/ngx_http_upstream.c[line2448]:ngx_http_upstream_process_non_buffered_request(), 
> a timer is added;
> 2. after that timer timeout, in 
> src/http/ngx_http_request.c[line2179]:ngx_http_set_write_handler(), 
> another timer is added again.
> 
> so the send_timeout is doubled.

Timeout is re-set on each send operation, and expected to be 
triggered when nothing happens during send_timeout.

Maxim Dounin


From pascal.nitsche at fansubcode.org  Wed May 18 18:35:04 2011
From: pascal.nitsche at fansubcode.org (Pascal Nitsche)
Date: Wed, 18 May 2011 16:35:04 +0200
Subject: ngx_http_secure_download report back to application?
In-Reply-To: <06d9bc75fdbcf1ed50b63659bdf8cc26.NginxMailingListEnglish@forum.nginx.org>
References: <4DD37A14.2050201@fansubcode.org>
	<06d9bc75fdbcf1ed50b63659bdf8cc26.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4DD3D918.8070800@fansubcode.org>

Hello Ensiferous,

thanks for your answer I think this will do the trick. Is there also a 
possibility to get the transfered bytes inside the request? (This would 
be really nice but is just optional.)

Thanks for your help again,

Pascal Nitsche

Am 18.05.2011 15:03, schrieb Ensiferous:
> Hi,
>
> You can do this with http://wiki.nginx.org/HttpCoreModule#post_action
> and http://wiki.nginx.org/HttpCoreModule#.24request_completion
>
> Sincerely,
> Martin Fjordvald
>
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198967,199037#msg-199037
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx



From nginx-forum at nginx.us  Wed May 18 20:11:33 2011
From: nginx-forum at nginx.us (myckeul)
Date: Wed, 18 May 2011 12:11:33 -0400
Subject: Proxy_pass and slash witch nginx
Message-ID: <ce02418eada188e33106a2a6e1b96b3f.NginxMailingListEnglish@forum.nginx.org>

Hello,

I set up a reverse proxy for a client with Nginx.

I still have a problem when I click on a link but the name of the
directory doesn't end with a slash.

The proxy address is http://proxy.com and all requests are returned to
http://mydomain.com

Example: http://proxy.com/test/ and reference works well inside
http://mydomain.com but remains on http://proxy.com/test/

If I remove the slash of the end I don't have http://proxy.com/test in
the address bar but http://mydomain.com/test/ (with slash at the end
because the other web server adds).

In the http headers I see a 301 but it is not handled by nginx visibly:

http://proxy.com/test


GET /test HTTP/1.1

Host: proxy.com
User-Agent: Mozilla/5.0 (X11; Linux i686; rv:2.0.1) Gecko/20100101
Firefox/4.0.1

Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8

Accept-Language: en-us,en;q=0.5

Accept-Encoding: gzip, deflate

Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7

Keep-Alive: 115

Connection: keep-alive

Cookie: language1=en; language2=en



HTTP/1.1 301 Moved Permanently

Server: nginx/0.7.67

Date: Wed, 18 May 2011 07:15:57 GMT

Content-Type: text/html; charset=iso-8859-1

Transfer-Encoding: chunked

Connection: keep-alive

Location: http://mydomain.com/test/

Age: 108

I thought to rewrite the rules but it might cause trouble.

If anyone has an idea on the subject, I'm interested.


Thank you

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198964,198964#msg-198964



From steven at deobald.ca  Wed May 18 23:35:49 2011
From: steven at deobald.ca (Steven Deobald)
Date: Wed, 18 May 2011 14:35:49 -0500
Subject: max_fails for 404s?
Message-ID: <BANLkTi=bQpPkC-n_XL4RBwK8rBAL6usnhQ@mail.gmail.com>

Hey folks,

We've got the following scenario occurring in production. The solution is
non-obvious... to me, at least.

[client] ==> [nginx] ==> [service S]

* nginx fronting 2 application servers which provide a service S. (Primary
and backup, both running Trinidad. Trinidad is a JRuby/Rails wrapper for
Tomcat.)
* hot-deploying a rails (jruby) app into Trinidad causes Trinidad to return
a 404 to nginx
* nginx returns the 404 to the application. In this particular case, the
client is another service which expects service S to remain live during
deployments

So, nginx does provide an "http_404" case for the "proxy_next_upstream"
directive. However, this would require the "max_fails" setting to pertain to
404s, which it doesn't... otherwise legitimate 404s produce an infinite
loop.

Is there something like "max_fails" for 404s?
Is there another solution to this problem?
Is it Trinidad's fault for returning 404s and not 503s? (I would say it is
but I can't find a solution to that problem just yet.)

Thanks!
-steven
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110518/4cea89de/attachment.html>

From mdounin at mdounin.ru  Wed May 18 23:50:45 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Wed, 18 May 2011 23:50:45 +0400
Subject: Proxy_pass and slash witch nginx
In-Reply-To: <ce02418eada188e33106a2a6e1b96b3f.NginxMailingListEnglish@forum.nginx.org>
References: <ce02418eada188e33106a2a6e1b96b3f.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110518195045.GX42265@mdounin.ru>

Hello!

On Wed, May 18, 2011 at 12:11:33PM -0400, myckeul wrote:

> Hello,
> 
> I set up a reverse proxy for a client with Nginx.
> 
> I still have a problem when I click on a link but the name of the
> directory doesn't end with a slash.
> 
> The proxy address is http://proxy.com and all requests are returned to
> http://mydomain.com
> 
> Example: http://proxy.com/test/ and reference works well inside
> http://mydomain.com but remains on http://proxy.com/test/
> 
> If I remove the slash of the end I don't have http://proxy.com/test in
> the address bar but http://mydomain.com/test/ (with slash at the end
> because the other web server adds).
> 
> In the http headers I see a 301 but it is not handled by nginx visibly:
> 
> http://proxy.com/test
> 
> 
> GET /test HTTP/1.1
> Host: proxy.com

[...]

> HTTP/1.1 301 Moved Permanently

[...]

> Location: http://mydomain.com/test/
> 
> Age: 108
> 
> I thought to rewrite the rules but it might cause trouble.
> 
> If anyone has an idea on the subject, I'm interested.

http://wiki.nginx.org/HttpProxyModule#proxy_redirect

Maxim Dounin


From antoine.bonavita at gmail.com  Wed May 18 23:53:39 2011
From: antoine.bonavita at gmail.com (Antoine Bonavita)
Date: Wed, 18 May 2011 21:53:39 +0200
Subject: max_fails for 404s?
In-Reply-To: <BANLkTi=bQpPkC-n_XL4RBwK8rBAL6usnhQ@mail.gmail.com>
References: <BANLkTi=bQpPkC-n_XL4RBwK8rBAL6usnhQ@mail.gmail.com>
Message-ID: <BANLkTimhGHsUq_M+y-jTuHnDM+7QDiV_ew@mail.gmail.com>

Hello,

>From my understanding you don't even want to set max_fails. From the
wiki (http://wiki.nginx.org/HttpUpstreamModule#server):
"max_fails = NUMBER - number of unsuccessful attempts at communicating
with the server within the time period (assigned by parameter
fail_timeout) after which it is considered inoperative"

While you hot-deploy server1 it will return 404. Because of
"proxy_next_upstream http_404", nginx will try server2 (which is
assumed to work). So, client sees response from server2.

When you're done with hot-deploying on server1, it should start
returning non-404 and you can safely hot-deploy server2.

Of course, as you are asking, this might be the configuration you
already have and it might not have this (expected) behavior. If you
have proxy_next_upstream http_404 in your configuration and it's not
working this way, you'll have to provide logs corresponding to the
hot-deployment if you want people here to give you a hand.

A.

--
Take a break. Visit nginx-land: http://www.nginx-discovery.com

On Wed, May 18, 2011 at 9:35 PM, Steven Deobald <steven at deobald.ca> wrote:
> Hey folks,
> We've got the following scenario?occurring?in production. The solution is
> non-obvious... to me, at least.
> [client] ==> [nginx] ==> [service S]
> * nginx fronting 2 application servers which provide a service S. (Primary
> and backup, both running Trinidad. Trinidad is a JRuby/Rails wrapper for
> Tomcat.)
> * hot-deploying a rails (jruby) app into Trinidad causes Trinidad to return
> a 404 to nginx
> * nginx returns the 404 to the application. In this particular case, the
> client is another service which expects service S to remain live during
> deployments
> So, nginx does provide an "http_404" case for the "proxy_next_upstream"
> directive. However, this would require the "max_fails" setting to pertain to
> 404s, which it doesn't... otherwise legitimate 404s produce an infinite
> loop.
> Is there something like "max_fails" for 404s?
> Is there another solution to this problem?
> Is it Trinidad's fault for returning 404s and not 503s? (I would say it is
> but I can't find a solution to that problem just yet.)
> Thanks!
> -steven
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>


From mdounin at mdounin.ru  Thu May 19 00:00:46 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 19 May 2011 00:00:46 +0400
Subject: max_fails for 404s?
In-Reply-To: <BANLkTi=bQpPkC-n_XL4RBwK8rBAL6usnhQ@mail.gmail.com>
References: <BANLkTi=bQpPkC-n_XL4RBwK8rBAL6usnhQ@mail.gmail.com>
Message-ID: <20110518200046.GY42265@mdounin.ru>

Hello!

On Wed, May 18, 2011 at 02:35:49PM -0500, Steven Deobald wrote:

> Hey folks,
> 
> We've got the following scenario occurring in production. The solution is
> non-obvious... to me, at least.
> 
> [client] ==> [nginx] ==> [service S]
> 
> * nginx fronting 2 application servers which provide a service S. (Primary
> and backup, both running Trinidad. Trinidad is a JRuby/Rails wrapper for
> Tomcat.)

Do you mean "backup" as in "server ... backup;" in upstream 
definition?

> * hot-deploying a rails (jruby) app into Trinidad causes Trinidad to return
> a 404 to nginx
> * nginx returns the 404 to the application. In this particular case, the
> client is another service which expects service S to remain live during
> deployments
> 
> So, nginx does provide an "http_404" case for the "proxy_next_upstream"
> directive. However, this would require the "max_fails" setting to pertain to
> 404s, which it doesn't... otherwise legitimate 404s produce an infinite
> loop.
> 
> Is there something like "max_fails" for 404s?
> Is there another solution to this problem?
> Is it Trinidad's fault for returning 404s and not 503s? (I would say it is
> but I can't find a solution to that problem just yet.)

Parameter max_fails is to mark backends down, and you probably 
don't want to mark backends down just because of some 
real/legitimate 404's.

On the other hand, using "server ... backup" indeed will cause 
infinite loop with "proxy_next_upstream http_404".  It's a bug, it 
should only ask each backend once.

Workaround is to don't use "server ... backup" but use either 
"weight=" instead (big one for real server, small one for backup 
one) or use error_page-based fallback instead.

Maxim Dounin


From nginx-forum at nginx.us  Thu May 19 00:20:31 2011
From: nginx-forum at nginx.us (mankins)
Date: Wed, 18 May 2011 16:20:31 -0400
Subject: proxy_cache 404 doesn't add headers
Message-ID: <a90803ea543b8c860dfb9c1b71cc0b1b.NginxMailingListEnglish@forum.nginx.org>

Hi.

In nginx 1.0.2 on Fedora, I'm using the proxy_cache in conjunction with
add_header to add the following header with cache-status (MISS,EXPIRED,
etc.) after the cache check:

add_header   X-Cache
"$upstream_cache_status,$upstream_status,$upstream_addr";

This works as expected for http status codes 200, 301, 302, but fails to
work for 404.

I've verified that the cache is actually working for 404 requests
(upstream sees no requests), but the extra add_header fails to get
added.

Any insight onto this behavior would be great. 

Thank you,
Matt Mankins

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199145,199145#msg-199145



From jelledejong at powercraft.nl  Thu May 19 00:28:16 2011
From: jelledejong at powercraft.nl (Jelle de Jong)
Date: Wed, 18 May 2011 22:28:16 +0200
Subject: searching example config for my setup
Message-ID: <4DD42BE0.7050705@powercraft.nl>

I am searching for an example configuration file for nginx for the
follow set-up:

- firewall with prerouting port 80 and 443 to server-x running nginx

- server-a running apache on port 80 (a.example.org) (192.168.24.60)
- server-b running apache on port 80 (b.example.org) (192.168.24.61)
- server-c running apache on port 80 (c.example.org) (192.168.24.62)
- ...
- all ip-based request should be return a not available error

How can I make a name-based redirection to the different servers? And
should I handle https on nginx or should apache handle it?

I am running Debian 6.0.1. with nginx 0.7.67

Thanks in advance,

Kind regards,

Jelle de Jong

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 316 bytes
Desc: OpenPGP digital signature
URL: <http://nginx.org/pipermail/nginx/attachments/20110518/3e566e66/attachment.pgp>

From steven at deobald.ca  Thu May 19 01:04:09 2011
From: steven at deobald.ca (Steven Deobald)
Date: Wed, 18 May 2011 16:04:09 -0500
Subject: max_fails for 404s?
In-Reply-To: <20110518200046.GY42265@mdounin.ru>
References: <BANLkTi=bQpPkC-n_XL4RBwK8rBAL6usnhQ@mail.gmail.com>
	<20110518200046.GY42265@mdounin.ru>
Message-ID: <BANLkTi=7u5mwcMEV775FQpgwFEhe6FmsuA@mail.gmail.com>

> Do you mean "backup" as in "server ... backup;" in upstream definition?

I do. And thank you for the detailed reply -- this was exactly the
information I was looking for!

Cheers,
-steven


On Wed, May 18, 2011 at 3:00 PM, Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
>
> On Wed, May 18, 2011 at 02:35:49PM -0500, Steven Deobald wrote:
>
> > Hey folks,
> >
> > We've got the following scenario occurring in production. The solution is
> > non-obvious... to me, at least.
> >
> > [client] ==> [nginx] ==> [service S]
> >
> > * nginx fronting 2 application servers which provide a service S.
> (Primary
> > and backup, both running Trinidad. Trinidad is a JRuby/Rails wrapper for
> > Tomcat.)
>
> Do you mean "backup" as in "server ... backup;" in upstream
> definition?
>
> > * hot-deploying a rails (jruby) app into Trinidad causes Trinidad to
> return
> > a 404 to nginx
> > * nginx returns the 404 to the application. In this particular case, the
> > client is another service which expects service S to remain live during
> > deployments
> >
> > So, nginx does provide an "http_404" case for the "proxy_next_upstream"
> > directive. However, this would require the "max_fails" setting to pertain
> to
> > 404s, which it doesn't... otherwise legitimate 404s produce an infinite
> > loop.
> >
> > Is there something like "max_fails" for 404s?
> > Is there another solution to this problem?
> > Is it Trinidad's fault for returning 404s and not 503s? (I would say it
> is
> > but I can't find a solution to that problem just yet.)
>
> Parameter max_fails is to mark backends down, and you probably
> don't want to mark backends down just because of some
> real/legitimate 404's.
>
> On the other hand, using "server ... backup" indeed will cause
> infinite loop with "proxy_next_upstream http_404".  It's a bug, it
> should only ask each backend once.
>
> Workaround is to don't use "server ... backup" but use either
> "weight=" instead (big one for real server, small one for backup
> one) or use error_page-based fallback instead.
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110518/2a67978d/attachment.html>

From jelledejong at powercraft.nl  Thu May 19 01:28:26 2011
From: jelledejong at powercraft.nl (Jelle de Jong)
Date: Wed, 18 May 2011 23:28:26 +0200
Subject: searching example config for my setup
In-Reply-To: <4DD42BE0.7050705@powercraft.nl>
References: <4DD42BE0.7050705@powercraft.nl>
Message-ID: <4DD439FA.2000705@powercraft.nl>

On 18-05-11 22:28, Jelle de Jong wrote:
> I am searching for an example configuration file for nginx for the
> follow set-up:

echo 'server {
    server_name a.example.org;
    location / {
        proxy_pass http://192.168.24.61/;
    }
}

server {
    server_name b.example.org;
    location / {
        proxy_pass http://192.168.24.62/;
    }
}

server {
    server_name c.example.org;
    location / {
        proxy_pass http://192.168.24.63/;
    }
}' | tee /etc/nginx/conf.d/02example.conf

Looking at several snippet I made the above config, it seems to partaly
do what I want... Is this the correct way (efficient/secure)?

Kind regards,

Jelle de Jong

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 316 bytes
Desc: OpenPGP digital signature
URL: <http://nginx.org/pipermail/nginx/attachments/20110518/4aee5a02/attachment.pgp>

From ngw at nofeed.org  Thu May 19 02:40:20 2011
From: ngw at nofeed.org (Nicholas Wieland)
Date: Wed, 18 May 2011 15:40:20 -0700
Subject: Serving websites from S3
Message-ID: <3B6D17C2-1812-4A83-B0B2-016E788356A0@nofeed.org>

Hi guys, I've tried to ask pretty much everywhere, but no solution until now. I'm quite sure I'm doing something wrong here, but to be honest I'm not able to catch it.

I'm rewriting urls from test.foo.com to bar.foo.com.s3.amazonaws.com/test, and everything works like a charm. If I point the browser to test.foo.com/index.html nginx serves my static website exactly as I want. The problem I'm facing now is that I'm not able to have nginx serve my index.html by default, while retrieving test.foo.com/

I've tried both try_files and index so far, with no result. I've also tried to catch 404s and serve my index.html page (even though I guess this is a bad practice).

This is what I have right now: https://gist.github.com/9a1ca5095392fb083f05

Can someone help me? I'm a bit out of ideas.

What is really strange to me is that (real example) http://test.stylej.am/ answers with a 404, I really have no idea why try_files doesn't work... Maybe $url/index.html is not right and I should use another syntax?

  ngw
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110518/d5d6f740/attachment.html>

From nginx-forum at nginx.us  Thu May 19 04:40:15 2011
From: nginx-forum at nginx.us (runesoerensen)
Date: Wed, 18 May 2011 20:40:15 -0400
Subject: Multiple SSL enabled hosts causes nginx to reload slowly
Message-ID: <cd7bf60a02009104ccbb55d25a61a2f3.NginxMailingListEnglish@forum.nginx.org>

Hi,

I've set up a ssl_certificate and ssl_certificate_key directives on the
http level of my nginx configuration. The problem I'm facing is that
starting/reloading nginx is getting slower and slower as more hosts
(server directives) are added. The server has roughly 1000 SSL-enabled
hosts that inherits the ssl certificate directives.

I know nginx verifies the certificate when loading it, so I'm wondering
if nginx checks the certificate each time an SSL enabled host inherits
the shared certificate? If so, shouldn't it only check the certificate
once? According to
http://nginx.org/en/docs/http/configuring_https_servers.html it should
cause all hosts to inherit a single memory footprint, but this doesn't
seem to be the case.

If this is a bug in nginx it may be an idea to completely disable
certificate verification if possible?

Any advice is highly appreciated.

Best,
Rune

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199166,199166#msg-199166



From wubingzheng at 163.com  Thu May 19 06:39:43 2011
From: wubingzheng at 163.com (Wu Bingzheng)
Date: Thu, 19 May 2011 10:39:43 +0800 (CST)
Subject: 2 questions on send_timeout (maybe bugs)
In-Reply-To: <20110518133507.GV42265@mdounin.ru>
References: <20110518133507.GV42265@mdounin.ru>
	<10a6803e.1907c.130026300cb.Coremail.wubingzheng@163.com>
Message-ID: <54e4f427.3d5f.130061f7663.Coremail.wubingzheng@163.com>

Hi Maxim,

Thanks for your answer. But I am still confused.


At 2011-05-18 21:35:07?"Maxim Dounin" <mdounin at mdounin.ru> wrote:

>Hello!
>
>On Wed, May 18, 2011 at 05:15:00PM +0800, Wu Bingzheng wrote:
>
>> hi all,
>> 
>> 
>> I meet 2 question when using nginx 1.0.0.
>> 
>> 
>> Question1.
>> 
>> what happens if the downstream connection breaks, before all 
>> data sent to the client?
>> 
>> 1. get client's request
>> 2. set handler: r->read_event_handle=ngx_http_block_reading
>> 3. send data to client, and blocked ( 
>> write_filter.c:c->send_chain() returns non-NULL or non-error), 
>> and go back to epoll_wait
>> 4. HERE, the downstream breaks
>> 5. epoll detect the breaking, so epoll_wait() continues in 
>> ngx_epoll_process_events(), where the revents==EPOLL_IN | 
>> EPOLL_ERR | EPOLL_HUP. So we call rev->handler(), which calls 
>> r->read_event_handle, which dose nothing. AS A RESULT, the 
>> downstream breaking was not detected by nginx !!!
>
>Here write event handler will be called.  While standard 
>ngx_http_writer() is in use - it will call output filter chain and 
>detect error.  Else it's up to active handler to correctly handle 
>this, e.g. standard upstream modules (proxy/fastcgi/...) may 
>either terminate request immediately or wait for full upstream 
>response depending on various settings (proxy_ignore_client_abort, 
>proxy_cache/proxy_store enabled).
>

You said the write handler will be called. But if a connection is break, revents returned by epoll_wait() is 'EPOLL_IN |  EPOLL_ERR | EPOLL_HUP', and no EPOLL_OUT in revents. So I think the write-handler will NOT be called.
The FACT is, in my server, there are many many access logs whose time-taken are 40.001s. (send_timeout is 20s, and this is a music online server, so many people re-choose a song before the current song finishs, which means the downstream breaks before nginx sends all data).

I add 2 debug log in ngx_epoll_process_events(), and copy part of the debug log here.
In this log, fd8 is downstream and fd9 is upstream. Only the downstream is what we will focus on (so ignore the fd9 please).
You can see that, in line 4, nginx write 496B to downstream, and the return value is non-NULL(line 5), which means the downstream is delayed.
In line 20, epoll_wait() returns with revents=25, which means the fd8(downstream) is *break*. But only a read-handler is called in line 21, and NO write-handler is called.
So, nginx doesnot detect the break of downstream, and waits for send_timeout, and *wakes* in line 32 and 35.

      0 ......
      1 2011/05/19 10:00:07 [debug] 26851#0: *1 write new buf t:0 f:0 0000000000000000, pos 00000000005AFB70, size: 4096 file: 0, size: 0
      2 2011/05/19 10:00:07 [debug] 26851#0: *1 http write filter: l:0 f:1 s:4096
      3 2011/05/19 10:00:07 [debug] 26851#0: *1 http write filter limit 0
      4 2011/05/19 10:00:07 [debug] 26851#0: *1 writev: 496
      5 2011/05/19 10:00:07 [debug] 26851#0: *1 http write filter 00000000005D91C0 
      6 2011/05/19 10:00:07 [debug] 26851#0: *1 copy filter: -2 "/data/music/885720/??~J???.mp3?xcode=efb9e826d9161812ab9483faf71dddbb"
      7 2011/05/19 10:00:07 [debug] 26851#0: *1 event timer add: 8: 20000:1305770427453
      8 2011/05/19 10:00:07 [debug] 26851#0: *1 event timer del: 9: 1305770467453
      9 2011/05/19 10:00:07 [debug] 26851#0: *1 epoll call wriet handler. fd: 9, revnets:5 
     10 2011/05/19 10:00:07 [debug] 26851#0: *1 http upstream request: "/data/music/885720/??~J???.mp3?xcode=efb9e826d9161812ab9483faf71dddbb"
     11 2011/05/19 10:00:07 [debug] 26851#0: *1 http upstream dummy handler
     12 2011/05/19 10:00:07 [debug] 26851#0: *1 epoll_wait: fd 9, revnets:5
     13 2011/05/19 10:00:07 [debug] 26851#0: *1 epoll call read handler. fd: 9, revnets:5
     14 2011/05/19 10:00:07 [debug] 26851#0: *1 http upstream request: "/data/music/885720/??~J???.mp3?xcode=efb9e826d9161812ab9483faf71dddbb"
     15 2011/05/19 10:00:07 [debug] 26851#0: *1 http upstream process non buffered upstream
     16 2011/05/19 10:00:07 [debug] 26851#0: *1 event timer: 8, old: 1305770427453, new: 1305770427453
     17 2011/05/19 10:00:07 [debug] 26851#0: *1 epoll call wriet handler. fd: 9, revnets:5
     18 2011/05/19 10:00:07 [debug] 26851#0: *1 http upstream request: "/data/music/885720/??~J???.mp3?xcode=efb9e826d9161812ab9483faf71dddbb"
     19 2011/05/19 10:00:07 [debug] 26851#0: *1 http upstream dummy handler
     20 2011/05/19 10:00:07 [debug] 26851#0: *1 epoll_wait: fd 8, revnets:25
     21 2011/05/19 10:00:07 [debug] 26851#0: *1 epoll call read handler. fd: 8, revnets:25
     22 2011/05/19 10:00:07 [debug] 26851#0: *1 http run request: "/data/music/885720/??~J???.mp3?xcode=efb9e826d9161812ab9483faf71dddbb"
     23 2011/05/19 10:00:07 [debug] 26851#0: *1 http reading blocked
     24 2011/05/19 10:00:07 [debug] 26851#0: *1 epoll_wait: fd 9, revnets:5
     25 2011/05/19 10:00:07 [debug] 26851#0: *1 epoll call read handler. fd: 9, revnets:5
     26 2011/05/19 10:00:07 [debug] 26851#0: *1 http upstream request: "/data/music/885720/??~J???.mp3?xcode=efb9e826d9161812ab9483faf71dddbb"
     27 2011/05/19 10:00:07 [debug] 26851#0: *1 http upstream process non buffered upstream
     28 2011/05/19 10:00:07 [debug] 26851#0: *1 event timer: 8, old: 1305770427453, new: 1305770427654
     29 2011/05/19 10:00:07 [debug] 26851#0: *1 epoll call wriet handler. fd: 9, revnets:5
     30 2011/05/19 10:00:07 [debug] 26851#0: *1 http upstream request: "/data/music/885720/??~J???.mp3?xcode=efb9e826d9161812ab9483faf71dddbb"
     31 2011/05/19 10:00:07 [debug] 26851#0: *1 http upstream dummy handler
     32 2011/05/19 10:00:27 [debug] 26851#0: *1 event timer del: 8: 1305770427453
     33 2011/05/19 10:00:27 [debug] 26851#0: *1 http run request: "/data/music/885720/??~J???.mp3?xcode=efb9e826d9161812ab9483faf71dddbb"
     34 2011/05/19 10:00:27 [debug] 26851#0: *1 http upstream process non buffered downstream
     35 2011/05/19 10:00:27 [info] 26851#0: *1 client timed out (110: Connection timed out) while sending to client, client: 61.135.169.175, server: z
        iyouziyuan.baidu.com, request: "GET /data/music/885720/%E4%BB%8A%E5%A4%A9.mp3?xcode=efb9e826d9161812ab9483faf71dddbb HTTP/1.0", upstream: "mem
        cached://172.18.254.50:11200"
     36 2011/05/19 10:00:27 [debug] 26851#0: *1 finalize http upstream request: 0
     37 2011/05/19 10:00:27 [debug] 26851#0: *1 finalize http memcached request 


>> 6. Then, we have to wait for send_timeout...
>> 
>> This situation dosenot occur everytime. If the downstream 
>> breaks, when nginx is sending data, the breaking will be 
>> detected.
>> But I can make this happen all the time, when:
>> 1. the client and server are far away from each other (e.g. not 
>> in a local net)
>> 2. using memc module
>>
>> Question2.
>> 
>> I set 'send_timeout' as 20s, but nginx timeouts in 40s actually.
>> 
>> 1. in 
>> src/http/ngx_http_upstream.c[line2448]:ngx_http_upstream_process_non_buffered_request(), 
>> a timer is added;
>> 2. after that timer timeout, in 
>> src/http/ngx_http_request.c[line2179]:ngx_http_set_write_handler(), 
>> another timer is added again.
>> 
>> so the send_timeout is doubled.
>
>Timeout is re-set on each send operation, and expected to be 
>triggered when nothing happens during send_timeout.

So, you think that, I set 'send_timeout' as 20s, but nginx actually timeout in 40s is right and expected?


>
>Maxim Dounin
>
>_______________________________________________
>nginx mailing list
>nginx at nginx.org
>http://nginx.org/mailman/listinfo/nginx



Thanks again for you responding.
Best Regards,
Wu


From nginx-forum at nginx.us  Thu May 19 06:52:39 2011
From: nginx-forum at nginx.us (ltss)
Date: Wed, 18 May 2011 22:52:39 -0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <c0c6df7135604bcee44a098e5b896a30@ruby-forum.com>
References: <c0c6df7135604bcee44a098e5b896a30@ruby-forum.com>
Message-ID: <e00b353ec548b2c91b169c5b8ccaaa9c.NginxMailingListEnglish@forum.nginx.org>

Hello i am new to nginx and i need some help to convert my .htaccess, i
have tried some shit but that not work ...

	Options FollowSymLinks
	RewriteEngine On

	RewriteRule ^(.*)\.php/img/(.*)$ img/$2 [C]
	RewriteRule ^/img/(.*)$ img/$2 [QSA,L]

	RewriteCond %{REQUEST_FILENAME} !^scriptsFtp.*$

	RewriteCond %{REQUEST_FILENAME} !-d
	RewriteCond %{REQUEST_FILENAME} !-f
		RewriteRule ^(.*)$ index.php [QSA,L]

Thanks

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,23868,199175#msg-199175



From justin at specialbusservice.com  Thu May 19 13:16:34 2011
From: justin at specialbusservice.com (Justin Cormack)
Date: Thu, 19 May 2011 10:16:34 +0100
Subject: Serving websites from S3
In-Reply-To: <3B6D17C2-1812-4A83-B0B2-016E788356A0@nofeed.org>
References: <3B6D17C2-1812-4A83-B0B2-016E788356A0@nofeed.org>
Message-ID: <1305796594.3761.6.camel@mackerel>

On Wed, 2011-05-18 at 15:40 -0700, Nicholas Wieland wrote:
> Hi guys, I've tried to ask pretty much everywhere, but no solution
> until now. I'm quite sure I'm doing something wrong here, but to be
> honest I'm not able to catch it.
> 
> 
> I'm rewriting urls from test.foo.com to
> bar.foo.com.s3.amazonaws.com/test, and everything works like a charm.
> If I point the browser to test.foo.com/index.html nginx serves my
> static website exactly as I want. The problem I'm facing now is that
> I'm not able to have nginx serve my index.html by default, while
> retrieving test.foo.com/
> 
> 
> I've tried both try_files and index so far, with no result. I've also
> tried to catch 404s and serve my index.html page (even though I guess
> this is a bad practice).
> 
> 
> This is what I have right
> now: https://gist.github.com/9a1ca5095392fb083f05
> 
> 
> Can someone help me? I'm a bit out of ideas.
> 
> 
> What is really strange to me is that (real example)
> http://test.stylej.am/ answers with a 404, I really have no idea why
> try_files doesn't work... Maybe $url/index.html is not right and I
> should use another syntax?

You know that Amazon added support for index files to S3?
http://aws.typepad.com/aws/2011/02/host-your-static-website-on-amazon-s3.html

Try_files does not work because it looks for acrual files on your file
system, and thats not where S3 is! You can use two locations, one to
match just / and one to match all other files.

Justin






From mdounin at mdounin.ru  Thu May 19 14:42:39 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 19 May 2011 14:42:39 +0400
Subject: Multiple SSL enabled hosts causes nginx to reload slowly
In-Reply-To: <cd7bf60a02009104ccbb55d25a61a2f3.NginxMailingListEnglish@forum.nginx.org>
References: <cd7bf60a02009104ccbb55d25a61a2f3.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110519104239.GZ42265@mdounin.ru>

Hello!

On Wed, May 18, 2011 at 08:40:15PM -0400, runesoerensen wrote:

> Hi,
> 
> I've set up a ssl_certificate and ssl_certificate_key directives on the
> http level of my nginx configuration. The problem I'm facing is that
> starting/reloading nginx is getting slower and slower as more hosts
> (server directives) are added. The server has roughly 1000 SSL-enabled
> hosts that inherits the ssl certificate directives.
> 
> I know nginx verifies the certificate when loading it, so I'm wondering
> if nginx checks the certificate each time an SSL enabled host inherits
> the shared certificate? If so, shouldn't it only check the certificate
> once? According to
> http://nginx.org/en/docs/http/configuring_https_servers.html it should
> cause all hosts to inherit a single memory footprint, but this doesn't
> seem to be the case.
> 
> If this is a bug in nginx it may be an idea to completely disable
> certificate verification if possible?
> 
> Any advice is highly appreciated.

Quick look suggests that time is taken mostly here in 
ngx_http_ssl_module.c:

    /* a temporary 512-bit RSA key is required for export versions of MSIE */
    if (ngx_ssl_generate_rsa512_key(&conf->ssl) != NGX_OK) {
        return NGX_CONF_ERROR;
    }

As export versions of MSIE are rare and hardly supported nowadays 
anyway, you may want to just nuke this code as a workaround.

I'll take a look at this further as time permits.

Maxim Dounin


From mdounin at mdounin.ru  Thu May 19 15:38:42 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 19 May 2011 15:38:42 +0400
Subject: 2 questions on send_timeout (maybe bugs)
In-Reply-To: <54e4f427.3d5f.130061f7663.Coremail.wubingzheng@163.com>
References: <20110518133507.GV42265@mdounin.ru>
	<10a6803e.1907c.130026300cb.Coremail.wubingzheng@163.com>
	<54e4f427.3d5f.130061f7663.Coremail.wubingzheng@163.com>
Message-ID: <20110519113842.GA42265@mdounin.ru>

Hello!

On Thu, May 19, 2011 at 10:39:43AM +0800, Wu Bingzheng wrote:

> Hi Maxim,
> 
> Thanks for your answer. But I am still confused.
> 
> 
> At 2011-05-18 21:35:07?"Maxim Dounin" <mdounin at mdounin.ru> wrote:
> 
> >Hello!
> >
> >On Wed, May 18, 2011 at 05:15:00PM +0800, Wu Bingzheng wrote:
> >
> >> hi all,
> >> 
> >> 
> >> I meet 2 question when using nginx 1.0.0.
> >> 
> >> 
> >> Question1.
> >> 
> >> what happens if the downstream connection breaks, before all 
> >> data sent to the client?
> >> 
> >> 1. get client's request
> >> 2. set handler: r->read_event_handle=ngx_http_block_reading
> >> 3. send data to client, and blocked ( 
> >> write_filter.c:c->send_chain() returns non-NULL or non-error), 
> >> and go back to epoll_wait
> >> 4. HERE, the downstream breaks
> >> 5. epoll detect the breaking, so epoll_wait() continues in 
> >> ngx_epoll_process_events(), where the revents==EPOLL_IN | 
> >> EPOLL_ERR | EPOLL_HUP. So we call rev->handler(), which calls 
> >> r->read_event_handle, which dose nothing. AS A RESULT, the 
> >> downstream breaking was not detected by nginx !!!
> >
> >Here write event handler will be called.  While standard 
> >ngx_http_writer() is in use - it will call output filter chain and 
> >detect error.  Else it's up to active handler to correctly handle 
> >this, e.g. standard upstream modules (proxy/fastcgi/...) may 
> >either terminate request immediately or wait for full upstream 
> >response depending on various settings (proxy_ignore_client_abort, 
> >proxy_cache/proxy_store enabled).
> >
> 
> You said the write handler will be called. But if a connection 
> is break, revents returned by epoll_wait() is 'EPOLL_IN |  
> EPOLL_ERR | EPOLL_HUP', and no EPOLL_OUT in revents. So I think 
> the write-handler will NOT be called.

Ah, ok, it looks like I was wrong here.  Write handler will be 
only called if there are no EPOLLIN flag in revents, and this is 
probably wrong assuming EPOLLHUP follows the same semantics as 
POLLHUP[1], i.e.

POLLHUP
...  This event and POLLOUT are mutually-exclusive; a stream can 
never be writable if a hangup has occurred. ...

Could you please provide detailed instructions to reproduce the 
problem, preferably with standard modules involved?

[1] http://pubs.opengroup.org/onlinepubs/9699919799/functions/poll.html

> The FACT is, in my server, there are many many access logs whose 
> time-taken are 40.001s. (send_timeout is 20s, and this is a 
> music online server, so many people re-choose a song before the 
> current song finishs, which means the downstream breaks before 
> nginx sends all data).
> 
> I add 2 debug log in ngx_epoll_process_events(), and copy part 
> of the debug log here.

Just a side note: mostly same information may be seen at global 
level debug log.

[...]

> >> I set 'send_timeout' as 20s, but nginx timeouts in 40s actually.
> >> 
> >> 1. in 
> >> src/http/ngx_http_upstream.c[line2448]:ngx_http_upstream_process_non_buffered_request(), 
> >> a timer is added;
> >> 2. after that timer timeout, in 
> >> src/http/ngx_http_request.c[line2179]:ngx_http_set_write_handler(), 
> >> another timer is added again.
> >> 
> >> so the send_timeout is doubled.
> >
> >Timeout is re-set on each send operation, and expected to be 
> >triggered when nothing happens during send_timeout.
> 
> So, you think that, I set 'send_timeout' as 20s, but nginx 
> actually timeout in 40s is right and expected?

It should timeout in 20s after last send.  Whether request time 
of 40s is wrong or not depends on many factors, but the code you 
pointed out looks fine.

Note also that in debug log you provided nginx correctly times out 
after 20s from last send.

Maxim Dounin


From jelledejong at powercraft.nl  Thu May 19 16:29:41 2011
From: jelledejong at powercraft.nl (Jelle de Jong)
Date: Thu, 19 May 2011 14:29:41 +0200
Subject: searching example config for my setup
In-Reply-To: <4DD439FA.2000705@powercraft.nl>
References: <4DD42BE0.7050705@powercraft.nl> <4DD439FA.2000705@powercraft.nl>
Message-ID: <4DD50D35.5070801@powercraft.nl>

On 18-05-11 23:28, Jelle de Jong wrote:
> On 18-05-11 22:28, Jelle de Jong wrote:
>> I am searching for an example configuration file for nginx for the
>> follow set-up:
> Looking at several snippet I made the above config, it seems to partaly
> do what I want... Is this the correct way (efficient/secure)?

server {
    listen      80;
    server_name a.powercraft.nl;
    access_log  /var/log/nginx/a.powercraft.nl.access.log;
    location / {
        proxy_set_header    Host $host;
        proxy_pass          http://192.168.24.63/;
    }
}

server {
    listen      80;
    server_name b.powercraft.nl;
    access_log  /var/log/nginx/c.powercraft.nl.access.log;
    location / {
        proxy_set_header    Host $host;
        proxy_pass          http://192.168.24.63/;
    }
}

server {
    listen      80;
    listen      443;
    server_name c.powercraft.nl;
    access_log  /var/log/nginx/c.powercraft.nl.access.log;
    location / {
        proxy_set_header    Host $host;
        proxy_pass          http://192.168.24.64/;
    }
}

I did some more research and found:
http://wiki.nginx.org/HttpProxyModule
http://wiki.nginx.org/VirtualHostExample

I came up with the above set-up that partly fills my need :)

Is it possbible to for http://192.168.24.63/ to see the IP-address of
the original browser requesting the http data? Now only the IP of the
nginx system is logged? (I know this is because of the proxy so I am
wondering if this is something that can be fixed?)

Is the configuration correct? Can it be improved? Hints & tips?

Kind regards,

Jelle de Jong

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 316 bytes
Desc: OpenPGP digital signature
URL: <http://nginx.org/pipermail/nginx/attachments/20110519/b1babf37/attachment.pgp>

From antoine.bonavita at gmail.com  Thu May 19 16:39:36 2011
From: antoine.bonavita at gmail.com (Antoine Bonavita)
Date: Thu, 19 May 2011 14:39:36 +0200
Subject: searching example config for my setup
In-Reply-To: <4DD50D35.5070801@powercraft.nl>
References: <4DD42BE0.7050705@powercraft.nl> <4DD439FA.2000705@powercraft.nl>
	<4DD50D35.5070801@powercraft.nl>
Message-ID: <BANLkTimk2yoXr_nUV6ZWYB5LxRx3W8JZRg@mail.gmail.com>

Hi,

Did you try "proxy_set_header  X-Real-IP  $remote_addr;" ?

A.

--
Take a break. Visit nginx-land: http://www.nginx-discovery.com ;)


On Thu, May 19, 2011 at 2:29 PM, Jelle de Jong
<jelledejong at powercraft.nl> wrote:
> On 18-05-11 23:28, Jelle de Jong wrote:
>> On 18-05-11 22:28, Jelle de Jong wrote:
>>> I am searching for an example configuration file for nginx for the
>>> follow set-up:
>> Looking at several snippet I made the above config, it seems to partaly
>> do what I want... Is this the correct way (efficient/secure)?
>
> server {
> ? ?listen ? ? ?80;
> ? ?server_name a.powercraft.nl;
> ? ?access_log ?/var/log/nginx/a.powercraft.nl.access.log;
> ? ?location / {
> ? ? ? ?proxy_set_header ? ?Host $host;
> ? ? ? ?proxy_pass ? ? ? ? ?http://192.168.24.63/;
> ? ?}
> }
>
> server {
> ? ?listen ? ? ?80;
> ? ?server_name b.powercraft.nl;
> ? ?access_log ?/var/log/nginx/c.powercraft.nl.access.log;
> ? ?location / {
> ? ? ? ?proxy_set_header ? ?Host $host;
> ? ? ? ?proxy_pass ? ? ? ? ?http://192.168.24.63/;
> ? ?}
> }
>
> server {
> ? ?listen ? ? ?80;
> ? ?listen ? ? ?443;
> ? ?server_name c.powercraft.nl;
> ? ?access_log ?/var/log/nginx/c.powercraft.nl.access.log;
> ? ?location / {
> ? ? ? ?proxy_set_header ? ?Host $host;
> ? ? ? ?proxy_pass ? ? ? ? ?http://192.168.24.64/;
> ? ?}
> }
>
> I did some more research and found:
> http://wiki.nginx.org/HttpProxyModule
> http://wiki.nginx.org/VirtualHostExample
>
> I came up with the above set-up that partly fills my need :)
>
> Is it possbible to for http://192.168.24.63/ to see the IP-address of
> the original browser requesting the http data? Now only the IP of the
> nginx system is logged? (I know this is because of the proxy so I am
> wondering if this is something that can be fixed?)
>
> Is the configuration correct? Can it be improved? Hints & tips?
>
> Kind regards,
>
> Jelle de Jong
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>


From nginx-forum at nginx.us  Thu May 19 17:20:55 2011
From: nginx-forum at nginx.us (TECK)
Date: Thu, 19 May 2011 09:20:55 -0400
Subject: Rewrite or internal redirection cycle?
Message-ID: <0843318a97adce238d8a671167f03ead.NginxMailingListEnglish@forum.nginx.org>

Hi,

I'm having a bit of an issue with the rewrite scheme:

server {
...
	location / {
		try_files		$uri $uri/ @data;
	}

	location @data {
		rewrite ^		/data.php$is_args;
		internal;
	}
...
}

When I access this request:
http://domain.com/information/feedback/?order=desc&sort=date

I get an internal redirection cycle:
*1 rewrite or internal redirection cycle while processing "/data.php?",
client: IP, server: domain.com, request: "GET
/information/feedback/?order=desc&sort=date HTTP/1.1", host:
"domain.com", referrer: "http://domain.com/information/feedback/"

However, if I access this request:
http://domain.com/information/feedback/400028-some-information.html

Everything works properly. What do you recommend to do, in order to
troubleshoot the issue?
My goal is to actually log what goes wrong with the php code, when the
redirection cycle is encountered.

Thanks for your help.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199260,199260#msg-199260



From malayter at gmail.com  Thu May 19 17:24:12 2011
From: malayter at gmail.com (Ryan Malayter)
Date: Thu, 19 May 2011 08:24:12 -0500
Subject: Virtualization
In-Reply-To: <BANLkTikHKd5PmbX6PzFduZ7ocizcWy3cgw@mail.gmail.com>
References: <BANLkTikHKd5PmbX6PzFduZ7ocizcWy3cgw@mail.gmail.com>
Message-ID: <BANLkTi=rBC6zKcjrmUX5mkQ6WsOCagRfwQ@mail.gmail.com>

On Fri, May 6, 2011 at 6:01 AM, Jonathan Leibiusky <ionathan at gmail.com> wrote:
> Hey!
> I am trying to virtualize nginx in my own private cloud. Is there any
> special consideration regard configuration that I have to consider?
> Anyone did something similar? Regard throughput, anyone with some
> numbers to share?
>

We run nginx on single-core VMs both internally and in our hosted
environment. Performance is excellent. WE allocate just 1-2 GB RAM per
vm. The stack is:
nginx 0.7.67->Ubuntu 10.04.2->Vmware ESX 4.1->Dell 2950III (Xeon 5300 series).

I recall seeing >8000  requests/sec using apache bench for a static
file 50K. And something like 5000 requests/sec for the same file
served through nginx being proxy_passed to an IIS backend (without
nginx proxy_cache). This is on a local LAN.

In our production internet-facing environment, which uses only
proxy_pass and proxy_cache in front of a Tomcat cluster, we routinely
see peaks 1000+ requests/sec through any single nginx VM. About 70% of
these requests are for static files that are usually hits in
proxy_cache, the rest are dynamic pages from the Tomcat servers. In
short, it just works. The only nginx tuning we really had to do was
adjusting proxy_buffers upwards, adding more workers, and setting more
connections per worker. We use 10 worker processes per core to get
around the lack of AIO on Linux.

As with all performance numbers, YMMV. There are*no* magic solutions,
and no substitutes for doing your own load testing on your own code
and hardware. We spent about 6 man-weeks building our load tests with
NeoLoad, and now we can re-run whenever we have a new code release or
are testing different infrastructure pieces.
-- 
RPM


From jelledejong at powercraft.nl  Thu May 19 17:46:26 2011
From: jelledejong at powercraft.nl (Jelle de Jong)
Date: Thu, 19 May 2011 15:46:26 +0200
Subject: searching example config for my setup
In-Reply-To: <BANLkTimk2yoXr_nUV6ZWYB5LxRx3W8JZRg@mail.gmail.com>
References: <4DD42BE0.7050705@powercraft.nl>
	<4DD439FA.2000705@powercraft.nl>	<4DD50D35.5070801@powercraft.nl>
	<BANLkTimk2yoXr_nUV6ZWYB5LxRx3W8JZRg@mail.gmail.com>
Message-ID: <4DD51F32.4010702@powercraft.nl>

Dear Antoine,

Thank you for taking the time to help.

On 19-05-11 14:39, Antoine Bonavita wrote:
> Did you try "proxy_set_header  X-Real-IP  $remote_addr;" ?

I did, but it sill shows the IP-addr of the nginx server....

server {
    listen      80;
    server_name sandbox01.powercraft.nl;
    access_log  /var/log/nginx/sandbox01.powercraft.nl.access.log;
    location / {
        proxy_pass          http://192.168.24.63/;
        proxy_set_header    X-Real-IP  $remote_addr;
        proxy_set_header    Host $host;
    }
}

# tail -n 1 /var/log/nginx/sandbox01.powercraft.nl.access.log;
109.32.215.197 - - [19/May/2011:15:42:41 +0200] "GET
/themes/garland/print.css?1 HTTP/1.1" 304 0
"http://sandbox01.powercraft.nl/admin/settings" "Mozilla/5.0 (X11; U;
Linux i686; en-US; rv:1.9.2.15) Gecko/20110303 Ubuntu/10.04 (lucid)
Firefox/3.6.15"

# tail -n 1 /var/log/apache2/sandbox01.powercraft.nl/access-default.log
192.168.24.62 - - [19/May/2011:15:42:41 +0200] "GET
/themes/garland/print.css?1 HTTP/1.0" 304 173
"http://sandbox01.powercraft.nl/admin/settings" "Mozilla/5.0 (X11; U;
Linux i686; en-US; rv:1.9.2.15) Gecko/20110303 Ubuntu/10.04 (lucid)
Firefox/3.6.15"

# 192.168.24.62 is the IP of the nginx server

Any ideas? Tips?

Kind regards,

Jelle de Jong

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 316 bytes
Desc: OpenPGP digital signature
URL: <http://nginx.org/pipermail/nginx/attachments/20110519/2393f927/attachment-0001.pgp>

From nginx-forum at nginx.us  Thu May 19 17:59:37 2011
From: nginx-forum at nginx.us (klausi)
Date: Thu, 19 May 2011 09:59:37 -0400
Subject: Disabling basic_auth with rewrites
Message-ID: <6a08a5933318a87afa453d03b955c65b.NginxMailingListEnglish@forum.nginx.org>

Hi,

I would like to generally protect my site but allow one certain path.
However, I need rewrites to forward the requested path to a PHP file.
Unfortunately the following does not work:

location ~ ^/feeds/importer/.* {
  auth_basic off;
  rewrite ^/(.*)$ /index.php?q=$1;
}

location ~ \.php$ {
  auth_basic "protected";
  auth_basic_user_file /etc/nginx/htpasswd/protected;
  fastcgi_split_path_info ^(.+\.php)(/.+)$;
  include fastcgi_params;
  fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
  fastcgi_intercept_errors on;
  fastcgi_pass $socket;
  # workaround as fastcgi_param cannot be used inside if statements
  set $https off;
  if ($scheme = https) {
    set $https on;
  }
  fastcgi_param HTTPS $https;
  fastcgi_read_timeout 6000;
}

location ~ ^/(.*) {
  auth_basic "protected";
  auth_basic_user_file /etc/nginx/htpasswd/protected;
  try_files $uri /index.php?q=$1&$args;
}

Any ideas how to unprotect the first location?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199278,199278#msg-199278



From mdounin at mdounin.ru  Thu May 19 18:09:50 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 19 May 2011 18:09:50 +0400
Subject: Disabling basic_auth with rewrites
In-Reply-To: <6a08a5933318a87afa453d03b955c65b.NginxMailingListEnglish@forum.nginx.org>
References: <6a08a5933318a87afa453d03b955c65b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110519140950.GB42265@mdounin.ru>

Hello!

On Thu, May 19, 2011 at 09:59:37AM -0400, klausi wrote:

> Hi,
> 
> I would like to generally protect my site but allow one certain path.
> However, I need rewrites to forward the requested path to a PHP file.
> Unfortunately the following does not work:
> 
> location ~ ^/feeds/importer/.* {
>   auth_basic off;
>   rewrite ^/(.*)$ /index.php?q=$1;
> }
> 
> location ~ \.php$ {
>   auth_basic "protected";
>   auth_basic_user_file /etc/nginx/htpasswd/protected;
>   fastcgi_split_path_info ^(.+\.php)(/.+)$;
>   include fastcgi_params;
>   fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
>   fastcgi_intercept_errors on;
>   fastcgi_pass $socket;
>   # workaround as fastcgi_param cannot be used inside if statements
>   set $https off;
>   if ($scheme = https) {
>     set $https on;
>   }
>   fastcgi_param HTTPS $https;
>   fastcgi_read_timeout 6000;
> }
> 
> location ~ ^/(.*) {
>   auth_basic "protected";
>   auth_basic_user_file /etc/nginx/htpasswd/protected;
>   try_files $uri /index.php?q=$1&$args;
> }
> 
> Any ideas how to unprotect the first location?

    location / {
        auth_basic "protected";
        auth_basic_user_file /etc/nginx/htpasswd/protected;
        ...

        location ~ \.php$ {
            fastcgi_pass ...
            ...
        }
    }

    location /feeds/importer/ {
        ...

        location ~ \.php$ {
            fastcgi_pass ...
            ...
        }
    }

Maxim Dounin


From antoine.bonavita at gmail.com  Thu May 19 19:35:17 2011
From: antoine.bonavita at gmail.com (Antoine Bonavita)
Date: Thu, 19 May 2011 17:35:17 +0200
Subject: searching example config for my setup
In-Reply-To: <4DD51F32.4010702@powercraft.nl>
References: <4DD42BE0.7050705@powercraft.nl> <4DD439FA.2000705@powercraft.nl>
	<4DD50D35.5070801@powercraft.nl>
	<BANLkTimk2yoXr_nUV6ZWYB5LxRx3W8JZRg@mail.gmail.com>
	<4DD51F32.4010702@powercraft.nl>
Message-ID: <BANLkTin=8wOk4w1y6Jr4_AOUwr+tPQmYeg@mail.gmail.com>

Hello,

On Thu, May 19, 2011 at 3:46 PM, Jelle de Jong
<jelledejong at powercraft.nl> wrote:
> Dear Antoine,
>
> Thank you for taking the time to help.
>
> On 19-05-11 14:39, Antoine Bonavita wrote:
>> Did you try "proxy_set_header ?X-Real-IP ?$remote_addr;" ?
>
> I did, but it sill shows the IP-addr of the nginx server....
>
> server {
> ? ?listen ? ? ?80;
> ? ?server_name sandbox01.powercraft.nl;
> ? ?access_log ?/var/log/nginx/sandbox01.powercraft.nl.access.log;
> ? ?location / {
> ? ? ? ?proxy_pass ? ? ? ? ?http://192.168.24.63/;
> ? ? ? ?proxy_set_header ? ?X-Real-IP ?$remote_addr;
> ? ? ? ?proxy_set_header ? ?Host $host;
> ? ?}
> }
>
> # tail -n 1 /var/log/nginx/sandbox01.powercraft.nl.access.log;
> 109.32.215.197 - - [19/May/2011:15:42:41 +0200] "GET
> /themes/garland/print.css?1 HTTP/1.1" 304 0
> "http://sandbox01.powercraft.nl/admin/settings" "Mozilla/5.0 (X11; U;
> Linux i686; en-US; rv:1.9.2.15) Gecko/20110303 Ubuntu/10.04 (lucid)
> Firefox/3.6.15"
>
> # tail -n 1 /var/log/apache2/sandbox01.powercraft.nl/access-default.log
> 192.168.24.62 - - [19/May/2011:15:42:41 +0200] "GET
> /themes/garland/print.css?1 HTTP/1.0" 304 173
> "http://sandbox01.powercraft.nl/admin/settings" "Mozilla/5.0 (X11; U;
> Linux i686; en-US; rv:1.9.2.15) Gecko/20110303 Ubuntu/10.04 (lucid)
> Firefox/3.6.15"
>
> # 192.168.24.62 is the IP of the nginx server
>
> Any ideas? Tips?
Sorry, I was assuming you were doing something else with the IP in the
backend than logging. As far as I know  the default logging for Apache
uses variable %a which is always set to the other end of the socket
(sandbox01.powercraft.nl in your example). Maybe you could configure
your Apache logs to use %{X-Real-IP}i instead. There you should see
what you expect in the Apache logs.

Good luck with it,

A.

--
Take a break. Visit nginx-land: http://www.nginx-discovery.com ;)


>
> Kind regards,
>
> Jelle de Jong
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>


From nginx-forum at nginx.us  Thu May 19 20:43:03 2011
From: nginx-forum at nginx.us (klausi)
Date: Thu, 19 May 2011 12:43:03 -0400
Subject: Disabling basic_auth with rewrites
In-Reply-To: <20110519140950.GB42265@mdounin.ru>
References: <20110519140950.GB42265@mdounin.ru>
Message-ID: <9739b40539caebc9cbb3d3c9dc3cdecd.NginxMailingListEnglish@forum.nginx.org>

Maxim Dounin Wrote:
-------------------------------------------------------
> 
>     location / {
>         auth_basic "protected";
>         auth_basic_user_file
> /etc/nginx/htpasswd/protected;
>         ...
> 
>         location ~ \.php$ {
>             fastcgi_pass ...
>             ...
>         }
>     }
> 
>     location /feeds/importer/ {
>         ...
> 
>         location ~ \.php$ {
>             fastcgi_pass ...
>             ...
>         }
>     }

Thanks for the quick reply, nested locations are nice, but they do not
help in this special case. A request to /feeds/importer/* has to be
rewritten to /index.php?q=feeds/importer/* and that should not be
protected. Is unprotecting a path with a special query possible at all?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199278,199319#msg-199319



From mdounin at mdounin.ru  Thu May 19 22:10:59 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 19 May 2011 22:10:59 +0400
Subject: Disabling basic_auth with rewrites
In-Reply-To: <9739b40539caebc9cbb3d3c9dc3cdecd.NginxMailingListEnglish@forum.nginx.org>
References: <20110519140950.GB42265@mdounin.ru>
	<9739b40539caebc9cbb3d3c9dc3cdecd.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110519181059.GD42265@mdounin.ru>

Hello!

On Thu, May 19, 2011 at 12:43:03PM -0400, klausi wrote:

> Maxim Dounin Wrote:
> -------------------------------------------------------
> > 
> >     location / {
> >         auth_basic "protected";
> >         auth_basic_user_file
> > /etc/nginx/htpasswd/protected;
> >         ...
> > 
> >         location ~ \.php$ {
> >             fastcgi_pass ...
> >             ...
> >         }
> >     }
> > 
> >     location /feeds/importer/ {
> >         ...
> > 
> >         location ~ \.php$ {
> >             fastcgi_pass ...
> >             ...
> >         }
> >     }
> 
> Thanks for the quick reply, nested locations are nice, but they do not
> help in this special case. A request to /feeds/importer/* has to be
> rewritten to /index.php?q=feeds/importer/* and that should not be
> protected. Is unprotecting a path with a special query possible at all?

Ah, sorry, I missed you actually want /feeds/importer/... to be 
fully handled by index.php.  This makes configuration even 
simplier:

    location / {
        auth_basic ...
        ...

        location ~ \.php$ {
            fastcgi_pass ...
            ...
        }
    }

    location /feeds/importer/ {
        rewrite ^/(.*) /index.php?q=$1? break;

        fastcgi_pass ...
        ...
    }
 
Note that the only goal of rewrite is to properly change url while 
correctly escaping new arguments and stripping old ones (note 
trailing '?'), as you probably don't want to allow unauthenticated 
users to supply arbitrary arguments to your index.php.  Due to 
'break' request doesn't leave the location in question after 
rewrite and processed there.

Maxim Dounin


From Richard.Kearsley at m247.com  Thu May 19 22:15:42 2011
From: Richard.Kearsley at m247.com (Richard Kearsley)
Date: Thu, 19 May 2011 18:15:42 +0000
Subject: lua: ngx.location.capture / limit proxy_cache individual file size
Message-ID: <EFC5A19DBF3A2E489904733B467D990B124575@office-server2.m247.local>

Hi
Is it possible to do a version of ngx.location.capture that only does a HEAD request rather than GET?
I'm only interested in the headers returned, the body would make what I'm trying to do pointless :(

Also if anyone has another method/idea of how to do this:
I want to limit requests for files of size greater than 512000 bytes using proxy_cache and actively deny and not cache them
I tried some stuff with 'if' but it's evaluated before the upstream-content-length is set :(





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110519/4a7fd2f5/attachment.html>

From igor at sysoev.ru  Thu May 19 23:16:56 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Thu, 19 May 2011 23:16:56 +0400
Subject: Disabling basic_auth with rewrites
In-Reply-To: <20110519181059.GD42265@mdounin.ru>
References: <20110519140950.GB42265@mdounin.ru>
	<9739b40539caebc9cbb3d3c9dc3cdecd.NginxMailingListEnglish@forum.nginx.org>
	<20110519181059.GD42265@mdounin.ru>
Message-ID: <20110519191656.GA48682@sysoev.ru>

On Thu, May 19, 2011 at 10:10:59PM +0400, Maxim Dounin wrote:
> Hello!
> 
> On Thu, May 19, 2011 at 12:43:03PM -0400, klausi wrote:
> 
> > Maxim Dounin Wrote:
> > -------------------------------------------------------
> > > 
> > >     location / {
> > >         auth_basic "protected";
> > >         auth_basic_user_file
> > > /etc/nginx/htpasswd/protected;
> > >         ...
> > > 
> > >         location ~ \.php$ {
> > >             fastcgi_pass ...
> > >             ...
> > >         }
> > >     }
> > > 
> > >     location /feeds/importer/ {
> > >         ...
> > > 
> > >         location ~ \.php$ {
> > >             fastcgi_pass ...
> > >             ...
> > >         }
> > >     }
> > 
> > Thanks for the quick reply, nested locations are nice, but they do not
> > help in this special case. A request to /feeds/importer/* has to be
> > rewritten to /index.php?q=feeds/importer/* and that should not be
> > protected. Is unprotecting a path with a special query possible at all?
> 
> Ah, sorry, I missed you actually want /feeds/importer/... to be 
> fully handled by index.php.  This makes configuration even 
> simplier:
> 
>     location / {
>         auth_basic ...
>         ...
> 
>         location ~ \.php$ {
>             fastcgi_pass ...
>             ...
>         }
>     }
> 
>     location /feeds/importer/ {
>         rewrite ^/(.*) /index.php?q=$1? break;
> 
>         fastcgi_pass ...
>         ...
>     }
>  
> Note that the only goal of rewrite is to properly change url while 
> correctly escaping new arguments and stripping old ones (note 
> trailing '?'), as you probably don't want to allow unauthenticated 
> users to supply arbitrary arguments to your index.php.  Due to 
> 'break' request doesn't leave the location in question after 
> rewrite and processed there.

My suggestion is to not use rewrite at all:

     location /feeds/importer/ {
         location ~ ^/(.*) {
             fastcgi_pass    ...
             fastcgi_param   SCRIPT_FILENAME  /path/to/index.php;
             fastcgi_param   QUERY_STRING     q=$1;
             ...
         }


-- 
Igor Sysoev


From kyle.burton at gmail.com  Thu May 19 23:34:55 2011
From: kyle.burton at gmail.com (Kyle R. Burton)
Date: Thu, 19 May 2011 15:34:55 -0400
Subject: Question about using the load balancer
Message-ID: <BANLkTinwhvAVCtmWjvOE3PBtO2rowT3oRA@mail.gmail.com>

Hello,

I am using nginx as a load balancer in front of a tier of 4 web
servers in a configuration very similar to:

  http://wiki.nginx.org/LoadBalanceExample

I am doing rolling deployments of my application on the 4 web servers.
 During a deploy, I take 2 of my 4 boxes out of the load balancer by
switching to an alternative nginx configuration and running 'nginx
reload'.  This has so far worked well.

One thing I noticed is that when I do this, nginx seems to snap
(break) any connections that were open at the time of the reload.  Is
there a configuration setting that I could use to allow these
connections to complete gracefuly, but otherwise have all new
connections be handled according to the updated configuration?

Best Regards,

Kyle Burton

-- 
Twitter: @kyleburton
Blog: http://asymmetrical-view.com/
Fun: http://snapclean.me/


From nginx-forum at nginx.us  Thu May 19 23:51:11 2011
From: nginx-forum at nginx.us (omega8cc)
Date: Thu, 19 May 2011 15:51:11 -0400
Subject: [NGINX] cache process manager consume a lot of CPU at reload and
	has to be restarted
In-Reply-To: <BANLkTimiy-i0gg1e0nW9wN5rySrLwhkcsA@mail.gmail.com>
References: <BANLkTimiy-i0gg1e0nW9wN5rySrLwhkcsA@mail.gmail.com>
Message-ID: <5574016561bfd454ad04bd477e558e06.NginxMailingListEnglish@forum.nginx.org>

Unfortunately, it still happens also for 1.0.2 release.

I have seen it over 10 times on two different servers since upgrade to
1.0.0, then 1.0.1 and 1.0.2.

Not sure how to debug this, as it is random issue, yet causes downtime
until our auto-healing will kill all nginx processes and wait for load
to stabilize to start nginx again.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,187322,199347#msg-199347



From ionathan at gmail.com  Thu May 19 23:52:54 2011
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Thu, 19 May 2011 16:52:54 -0300
Subject: loadbalancing
Message-ID: <BANLkTinps27UqTX6qUpYN219eVV7N0+COQ@mail.gmail.com>

Hi!
It's been a while now that I am struggling with nginx round-robin load
balancer.
The problem that I am experiencing is that when a server goes down, the
entire upstream is not balanced anymore. The next server will get double
traffic. And if that one goes down, the next one will get 3 times the
traffice, and so on.

If I have something like this:

server localhost max_fails=2 fail_timeout=60s;

Does it mean that as soon as I get 2 fails within 60s it will mark the
server as down without waiting the entire 60s? or is it going to wait 60s to
mark the server down?

If it waits 60s, how can I use round-robin and make sure that when a server
goes down the rest will get traffic evenly?

Thanks,

Jonathan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110519/d88c29c3/attachment.html>

From nginx-forum at nginx.us  Fri May 20 00:34:34 2011
From: nginx-forum at nginx.us (nickn)
Date: Thu, 19 May 2011 16:34:34 -0400
Subject: Randomly stopping large file downloads?
Message-ID: <afcd4c4d19140b43df79d82ea9980087.NginxMailingListEnglish@forum.nginx.org>

Hi,
1.) What would cause nginx to randomly stop serving large file
downloads? When I try to download a 800MB file it stops serving the file
after between 10 and 30 MB. I've had to revert to lighttpd because of
this.

I checked the usual things: There wasn't anything in the error log, the
system had plenty of memory (several hundred MB free) and the worker
processes were all still there after the download dies. It doesn't
matter what browser I use.

Any idea what would cause this?



2.) BTW there's a bug in the forum.nginx.org forum where when you first
get an account it prepopulates the login form with your email address
rather than your user name. If you try to log in like that it won't let
you in. You have to clear out what it prepopulated and enter your
username. This is confusing.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199359,199359#msg-199359



From nginx-forum at nginx.us  Fri May 20 00:37:38 2011
From: nginx-forum at nginx.us (myckeul)
Date: Thu, 19 May 2011 16:37:38 -0400
Subject: Proxy_pass and slash witch nginx
In-Reply-To: <20110518195045.GX42265@mdounin.ru>
References: <20110518195045.GX42265@mdounin.ru>
Message-ID: <0cbde934384bd99d3e3d381c2346af82.NginxMailingListEnglish@forum.nginx.org>

Hello,

Thanks for your response but they won't working.

Here is my conf : 

nginx.conf :
user www-data;
worker_processes  1;

error_log  /var/log/nginx/error.log;
pid        /var/run/nginx.pid;

events {
    worker_connections  1024;
    # multi_accept on;
}

http {
    include       /etc/nginx/mime.types;

    access_log	/var/log/nginx/access.log;

    sendfile        on;
    #tcp_nopush     on;

    #keepalive_timeout  0;
    keepalive_timeout  65;
    tcp_nodelay        on;

    gzip  on;
    gzip_disable "MSIE [1-6]\.(?!.*SV1)";

    include /etc/nginx/conf.d/*.conf;
    include /etc/nginx/sites-enabled/*;
}

my vhost : 
server {
       listen [::]:80;

       server_name proxy.com *.proxy.com;

       access_log  /var/log/nginx/proxy_.access.log;

       error_page 404 = /404.htm;

       location /404.htm {
       	root /var/www;
       }

       location / {
       		proxy_set_header X-Forwarded-Host $host;
		proxy_set_header X-Forwarded-Server $host;
		proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;


		if ($host ~* ^proxy\.com$) {
		   proxy_pass http://mydomain.com;
		}

		if ($host ~* ^www\.proxy\.com$) {
		   proxy_pass http://www.mydomain.com;
		}

		if ($host ~* ^subdomain\.proxy\.com$) {
		   proxy_pass http://subdomain.mydomain.com;
		}

		proxy_redirect http://proxy.com/test http://proxy.com/test/;
	}
}

and my proxy.conf file : 
proxy_redirect          off;
proxy_intercept_errors  on;
proxy_set_header        Host            $host;
proxy_set_header        X-Real-IP       $remote_addr;
proxy_set_header        X-Forwarded-For $proxy_add_x_forwarded_for;
client_max_body_size    10m;
client_body_buffer_size 128k;
client_header_buffer_size 64k;
proxy_connect_timeout   90;
proxy_send_timeout      90;
proxy_read_timeout      90;
proxy_buffer_size   16k;
proxy_buffers       32   16k;
proxy_busy_buffers_size 64k;


Thanks for your help

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198964,199196#msg-199196



From nginx-forum at nginx.us  Fri May 20 00:54:48 2011
From: nginx-forum at nginx.us (omega8cc)
Date: Thu, 19 May 2011 16:54:48 -0400
Subject: [NGINX] cache process manager consume a lot of CPU at reload and
	has to be restarted
In-Reply-To: <5574016561bfd454ad04bd477e558e06.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTimiy-i0gg1e0nW9wN5rySrLwhkcsA@mail.gmail.com>
	<5574016561bfd454ad04bd477e558e06.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <202fe8694c21a8ab5aa0b60b79072cc2.NginxMailingListEnglish@forum.nginx.org>

One possible reason could be some race condition when the server is
still running in the reload state and it receives another reload command
- this can be caused by provision in Aegir system for Drupal, as it
reloads nginx sometimes two times during one action, like platform/site
migration to make sure all vhosts are reloaded correctly.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,187322,199366#msg-199366



From jim at ohlste.in  Fri May 20 01:01:01 2011
From: jim at ohlste.in (Jim Ohlstein)
Date: Thu, 19 May 2011 17:01:01 -0400
Subject: Randomly stopping large file downloads?
In-Reply-To: <afcd4c4d19140b43df79d82ea9980087.NginxMailingListEnglish@forum.nginx.org>
References: <afcd4c4d19140b43df79d82ea9980087.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4DD5850D.8010909@ohlste.in>

On 5/19/11 4:34 PM, nickn wrote:
> Hi,
> 1.) What would cause nginx to randomly stop serving large file
> downloads? When I try to download a 800MB file it stops serving the file
> after between 10 and 30 MB. I've had to revert to lighttpd because of
> this.
> 
> I checked the usual things: There wasn't anything in the error log, the
> system had plenty of memory (several hundred MB free) and the worker
> processes were all still there after the download dies. It doesn't
> matter what browser I use.
> 
> Any idea what would cause this?
> 
> 
> 
> 2.) BTW there's a bug in the forum.nginx.org forum where when you first
> get an account it prepopulates the login form with your email address
> rather than your user name. If you try to log in like that it won't let
> you in. You have to clear out what it prepopulated and enter your
> username. This is confusing.

First, this is an issue with how your browser is "pre-populating" that
field, not the forum software. See attached for what I see when I
attempt to log in. This behavior is consistent in Firefox, Chrome, and
Safari.

Second, a "Username" is a user name and an email is an email. You should
be able to distinguish between the two. Again, please see the attached
login form. This hardly seems "confusing" to me. Again please see the
attached.

-- 
Jim Ohlstein
-------------- next part --------------
A non-text attachment was scrubbed...
Name: login.png
Type: image/png
Size: 3680 bytes
Desc: not available
URL: <http://nginx.org/pipermail/nginx/attachments/20110519/780d8074/attachment.png>

From nginx-forum at nginx.us  Fri May 20 01:29:25 2011
From: nginx-forum at nginx.us (nickn)
Date: Thu, 19 May 2011 17:29:25 -0400
Subject: Randomly stopping large file downloads?
In-Reply-To: <4DD5850D.8010909@ohlste.in>
References: <4DD5850D.8010909@ohlste.in>
Message-ID: <3352d7c0910978a8a1b3a07e8afa2562.NginxMailingListEnglish@forum.nginx.org>

An ideas why nginx would be cutting the downloads short?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199359,199377#msg-199377



From Richard.Kearsley at m247.com  Fri May 20 01:59:16 2011
From: Richard.Kearsley at m247.com (Richard Kearsley)
Date: Thu, 19 May 2011 21:59:16 +0000
Subject: lua location capture
Message-ID: <EFC5A19DBF3A2E489904733B467D990B12493E@office-server2.m247.local>

Hi
The lua doc at https://github.com/chaoslawful/lua-nginx-module
Says "Note that, by default, subrequests issued by ngx.location.capture inherit all the request headers of the current request."
And goes on to say that it can be an unwanted feature
In my project I actually take advantage of this feature, but for some reason the headers don't go through to the sub request
All I get is:

Uri:/checksize/100meg.zip
Host : 127.0.0.1:12340
Connection : close
Content-Length : 0

It should have a TON of headers and I've proven they are there in the main request
Any ideas why they don't get through?

Thanks as always
Richard
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110519/2e91e55a/attachment.html>

From david at styleflare.com  Fri May 20 02:10:04 2011
From: david at styleflare.com (David J.)
Date: Thu, 19 May 2011 18:10:04 -0400
Subject: try_files
Message-ID: <4DD5953C.7030202@styleflare.com>

I have a uwsgi_app

I was wondering why

location / {
root /myapp;
index index.html index.wsgi;
try_files $uri $uri/ @wsgi;
}
;

When I visit the domain.com  returns a 403 (when I dont have an 
index.html file in the root directory);

but if I do domain.com/index.wsgi

The WSGI app is called. This all works.

I am guessing I am doing something wrong.

In my config I have

location @wsgi {
     include uwsgi_params;
     uwsgi_pass unix:/tmp/uwsgi.sock
}


From agentzh at gmail.com  Fri May 20 07:56:21 2011
From: agentzh at gmail.com (agentzh)
Date: Fri, 20 May 2011 11:56:21 +0800
Subject: lua location capture
In-Reply-To: <EFC5A19DBF3A2E489904733B467D990B12493E@office-server2.m247.local>
References: <EFC5A19DBF3A2E489904733B467D990B12493E@office-server2.m247.local>
Message-ID: <BANLkTi=_eN=EzKF-J9=cFgu6qd29GVOwuA@mail.gmail.com>

On Fri, May 20, 2011 at 5:59 AM, Richard Kearsley
<Richard.Kearsley at m247.com> wrote:
> Hi
>
> The lua doc at https://github.com/chaoslawful/lua-nginx-module
>
> Says ?Note that, by default, subrequests issued by ngx.location.capture
> inherit all the request headers of the current request.?
>
> And goes on to say that it can be an unwanted feature
>
> In my project I actually take advantage of this feature, but for some reason
> the headers don?t go through to the sub request
>

Could you please paste your nginx.conf here? Also, information about
your nginx's version and ngx_lua's version can also be helpful :)

Best,
-agentzh


From wubingzheng at 163.com  Fri May 20 07:58:20 2011
From: wubingzheng at 163.com (Wu Bingzheng)
Date: Fri, 20 May 2011 11:58:20 +0800 (CST)
Subject: 2 questions on send_timeout (maybe bugs)
In-Reply-To: <20110519113842.GA42265@mdounin.ru>
References: <20110519113842.GA42265@mdounin.ru>
	<20110518133507.GV42265@mdounin.ru>
	<10a6803e.1907c.130026300cb.Coremail.wubingzheng@163.com>
	<54e4f427.3d5f.130061f7663.Coremail.wubingzheng@163.com>
Message-ID: <76e2ab59.13e9e.1300b8dcee7.Coremail.wubingzheng@163.com>

Hi Maxim,

1.
I am so sorry that I made a big mistake on question 1.
Nginx didn't detect the downstream breaking, because I use "proxy_ignore_client_abort on;".
So Nginx works correctlly. I am so sorry about it.


2.
For question 2.
You said that :
>Note also that in debug log you provided nginx correctly times out 
>after 20s from last send.

That's because I set the send_timeout as 10s. I am sorry that I did not make clear of that.


You can reproduce the doubled timeout easy:

[server] using nginx-1.0.0, without any addon module. in nginx.conf:

    server {
        listen       5821;
        proxy_buffering off; #must in non-buffering mode
        send_timeout 10;

        location / {
            proxy_pass http://sourceserver;
        }
    }



[client] you can run this simple python script:

  #!/usr/bin/python
  import socket
  import time
  s = socket.socket(socket.AF_INET,socket.SOCK_STREAM)
  s.connect((127.0.0.1, 5821))
  s.send("GET /data/something HTTP/1.0\r\n\r\n")
  buf = s.recv(100)
  print buf
  time.sleep(100)



Then, in access.log, you will find that the request is finished in 20s.


Best Regads,
Wu








At 2011-05-19 19:38:42?"Maxim Dounin" <mdounin at mdounin.ru> wrote:

>Hello!
>
>On Thu, May 19, 2011 at 10:39:43AM +0800, Wu Bingzheng wrote:
>
>> Hi Maxim,
>> 
>> Thanks for your answer. But I am still confused.
>> 
>> 
>> At 2011-05-18 21:35:07?"Maxim Dounin" <mdounin at mdounin.ru> wrote:
>> 
>> >Hello!
>> >
>> >On Wed, May 18, 2011 at 05:15:00PM +0800, Wu Bingzheng wrote:
>> >
>> >> hi all,
>> >> 
>> >> 
>> >> I meet 2 question when using nginx 1.0.0.
>> >> 
>> >> 
>> >> Question1.
>> >> 
>> >> what happens if the downstream connection breaks, before all 
>> >> data sent to the client?
>> >> 
>> >> 1. get client's request
>> >> 2. set handler: r->read_event_handle=ngx_http_block_reading
>> >> 3. send data to client, and blocked ( 
>> >> write_filter.c:c->send_chain() returns non-NULL or non-error), 
>> >> and go back to epoll_wait
>> >> 4. HERE, the downstream breaks
>> >> 5. epoll detect the breaking, so epoll_wait() continues in 
>> >> ngx_epoll_process_events(), where the revents==EPOLL_IN | 
>> >> EPOLL_ERR | EPOLL_HUP. So we call rev->handler(), which calls 
>> >> r->read_event_handle, which dose nothing. AS A RESULT, the 
>> >> downstream breaking was not detected by nginx !!!
>> >
>> >Here write event handler will be called.  While standard 
>> >ngx_http_writer() is in use - it will call output filter chain and 
>> >detect error.  Else it's up to active handler to correctly handle 
>> >this, e.g. standard upstream modules (proxy/fastcgi/...) may 
>> >either terminate request immediately or wait for full upstream 
>> >response depending on various settings (proxy_ignore_client_abort, 
>> >proxy_cache/proxy_store enabled).
>> >
>> 
>> You said the write handler will be called. But if a connection 
>> is break, revents returned by epoll_wait() is 'EPOLL_IN |  
>> EPOLL_ERR | EPOLL_HUP', and no EPOLL_OUT in revents. So I think 
>> the write-handler will NOT be called.
>
>Ah, ok, it looks like I was wrong here.  Write handler will be 
>only called if there are no EPOLLIN flag in revents, and this is 
>probably wrong assuming EPOLLHUP follows the same semantics as 
>POLLHUP[1], i.e.
>
>POLLHUP
>...  This event and POLLOUT are mutually-exclusive; a stream can 
>never be writable if a hangup has occurred. ...
>
>Could you please provide detailed instructions to reproduce the 
>problem, preferably with standard modules involved?
>
>[1] http://pubs.opengroup.org/onlinepubs/9699919799/functions/poll.html
>
>> The FACT is, in my server, there are many many access logs whose 
>> time-taken are 40.001s. (send_timeout is 20s, and this is a 
>> music online server, so many people re-choose a song before the 
>> current song finishs, which means the downstream breaks before 
>> nginx sends all data).
>> 
>> I add 2 debug log in ngx_epoll_process_events(), and copy part 
>> of the debug log here.
>
>Just a side note: mostly same information may be seen at global 
>level debug log.
>
>[...]
>
>> >> I set 'send_timeout' as 20s, but nginx timeouts in 40s actually.
>> >> 
>> >> 1. in 
>> >> src/http/ngx_http_upstream.c[line2448]:ngx_http_upstream_process_non_buffered_request(), 
>> >> a timer is added;
>> >> 2. after that timer timeout, in 
>> >> src/http/ngx_http_request.c[line2179]:ngx_http_set_write_handler(), 
>> >> another timer is added again.
>> >> 
>> >> so the send_timeout is doubled.
>> >
>> >Timeout is re-set on each send operation, and expected to be 
>> >triggered when nothing happens during send_timeout.
>> 
>> So, you think that, I set 'send_timeout' as 20s, but nginx 
>> actually timeout in 40s is right and expected?
>
>It should timeout in 20s after last send.  Whether request time 
>of 40s is wrong or not depends on many factors, but the code you 
>pointed out looks fine.
>
>Note also that in debug log you provided nginx correctly times out 
>after 20s from last send.
>
>Maxim Dounin



From joshua.pearce at gmail.com  Fri May 20 05:54:41 2011
From: joshua.pearce at gmail.com (Joshua Pearce)
Date: Thu, 19 May 2011 21:54:41 -0400
Subject: How to Set Expires With Try Files
Message-ID: <5E68732E-D6D1-45DA-B2A5-85E65687FC72@gmail.com>

Below is my config file. How can I change the two "if" blocks to try_files, and still do a 30d expires for static files?

server {
        listen   80;
        server_name  www.joshpearce.org;
        client_max_body_size 4M;
        access_log  /var/log/nginx/localhost.access.log;

        # root of website *exactly*, use '=' for speed
        location = / {
                root /var/www/joshpearce.org;
                index index.php index.html index.htm;
        }
        location / {
                root   /var/www/joshpearce.org;
                index index.php index.html index.htm;
                if (-f $request_filename) {
                        expires 30d;
                        break;
                }
                if (!-e $request_filename) {
                        rewrite ^(.+)$ /index.php?q=$1 last;
                }
        }
        # use: htpasswd - htpwds NewUser NewPassword
        location ^~ /info {
                auth_basic  "Restricted";
                auth_basic_user_file /etc/nginx/htpwds;
                include /etc/nginx/fastcgi_params;
                fastcgi_pass  unix:/usr/local/var/run/php-fpm.sock;
                fastcgi_index index.php;
                fastcgi_param SCRIPT_FILENAME /var/www/joshpearce.org$fastcgi_script_name;
        }
        location ~ \.php$ {
                include /etc/nginx/fastcgi_params;
                fastcgi_pass  unix:/usr/local/var/run/php-fpm.sock;
                fastcgi_index index.php;
                fastcgi_param SCRIPT_FILENAME /var/www/joshpearce.org$fastcgi_script_name;
        }

        #error_page  404  /404.html;
}


From igor at sysoev.ru  Fri May 20 09:34:36 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 20 May 2011 09:34:36 +0400
Subject: Rewrite or internal redirection cycle?
In-Reply-To: <0843318a97adce238d8a671167f03ead.NginxMailingListEnglish@forum.nginx.org>
References: <0843318a97adce238d8a671167f03ead.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110520053436.GB73208@sysoev.ru>

On Thu, May 19, 2011 at 09:20:55AM -0400, TECK wrote:
> Hi,
> 
> I'm having a bit of an issue with the rewrite scheme:
> 
> server {
> ...
> 	location / {
> 		try_files		$uri $uri/ @data;
> 	}
> 
> 	location @data {
> 		rewrite ^		/data.php$is_args;
> 		internal;
> 	}
> ...
> }
> 
> When I access this request:
> http://domain.com/information/feedback/?order=desc&sort=date
> 
> I get an internal redirection cycle:
> *1 rewrite or internal redirection cycle while processing "/data.php?",
> client: IP, server: domain.com, request: "GET
> /information/feedback/?order=desc&sort=date HTTP/1.1", host:
> "domain.com", referrer: "http://domain.com/information/feedback/"
> 
> However, if I access this request:
> http://domain.com/information/feedback/400028-some-information.html
> 
> Everything works properly. What do you recommend to do, in order to
> troubleshoot the issue?
> My goal is to actually log what goes wrong with the php code, when the
> redirection cycle is encountered.
> 
> Thanks for your help.

You should either add "break" to stop internal redirection cycle:

	location @data {
		rewrite ^	/data.php$args  break;
	}

or should define /data.php in try_files

	location / {
		try_files	$uri $uri/ /index.php?$args;
	}

or (my favorite) should define exactly what nginx should do in @data:

	location @data {
		fastcgi_pass  ...
		fastcgi_param  SCRIPT_FILENAME  /path/to/data.php;
		...
	}

The later way looks more complex at the first sight, but it creates
independed locations and allows configuration to grow at very large scale.


-- 
Igor Sysoev


From igor at sysoev.ru  Fri May 20 09:36:15 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 20 May 2011 09:36:15 +0400
Subject: How to Set Expires With Try Files
In-Reply-To: <5E68732E-D6D1-45DA-B2A5-85E65687FC72@gmail.com>
References: <5E68732E-D6D1-45DA-B2A5-85E65687FC72@gmail.com>
Message-ID: <20110520053615.GC73208@sysoev.ru>

On Thu, May 19, 2011 at 09:54:41PM -0400, Joshua Pearce wrote:
> Below is my config file. How can I change the two "if" blocks to try_files, and still do a 30d expires for static files?
> 
> server {
>         listen   80;
>         server_name  www.joshpearce.org;
>         client_max_body_size 4M;
>         access_log  /var/log/nginx/localhost.access.log;
> 
>         # root of website *exactly*, use '=' for speed
>         location = / {
>                 root /var/www/joshpearce.org;
>                 index index.php index.html index.htm;
>         }
>         location / {
>                 root   /var/www/joshpearce.org;
>                 index index.php index.html index.htm;
>                 if (-f $request_filename) {
>                         expires 30d;
>                         break;
>                 }
>                 if (!-e $request_filename) {
>                         rewrite ^(.+)$ /index.php?q=$1 last;
>                 }
>         }
>         # use: htpasswd - htpwds NewUser NewPassword
>         location ^~ /info {
>                 auth_basic  "Restricted";
>                 auth_basic_user_file /etc/nginx/htpwds;
>                 include /etc/nginx/fastcgi_params;
>                 fastcgi_pass  unix:/usr/local/var/run/php-fpm.sock;
>                 fastcgi_index index.php;
>                 fastcgi_param SCRIPT_FILENAME /var/www/joshpearce.org$fastcgi_script_name;
>         }
>         location ~ \.php$ {
>                 include /etc/nginx/fastcgi_params;
>                 fastcgi_pass  unix:/usr/local/var/run/php-fpm.sock;
>                 fastcgi_index index.php;
>                 fastcgi_param SCRIPT_FILENAME /var/www/joshpearce.org$fastcgi_script_name;
>         }
> 
>         #error_page  404  /404.html;
> }

You do not need this block:

                 if (-f $request_filename) {
                         expires 30d;
                         break;
                 }

at all, since nginx adds Expires only to existant files.


-- 
Igor Sysoev


From agentzh at gmail.com  Fri May 20 10:51:40 2011
From: agentzh at gmail.com (agentzh)
Date: Fri, 20 May 2011 14:51:40 +0800
Subject: lua: ngx.location.capture / limit proxy_cache individual file size
In-Reply-To: <EFC5A19DBF3A2E489904733B467D990B124575@office-server2.m247.local>
References: <EFC5A19DBF3A2E489904733B467D990B124575@office-server2.m247.local>
Message-ID: <BANLkTimNUcE0KVY+kwAkHS4zB48x+G4nGQ@mail.gmail.com>

On Fri, May 20, 2011 at 2:15 AM, Richard Kearsley
<Richard.Kearsley at m247.com> wrote:
> Is it possible to do a version of ngx.location.capture that only does a HEAD
> request rather than GET?
>

Please try

    ngx.location.capture("/some-location", { method = ngx.HTTP_HEAD })

If it does not work, please consider file a bug ticket on GitHub:

    https://github.com/chaoslawful/lua-nginx-module/issues

> I?m only interested in the headers returned, the body would make what I?m
> trying to do pointless L
>

*nod*

Best,
-agentzh


From nginx-forum at nginx.us  Fri May 20 10:54:31 2011
From: nginx-forum at nginx.us (arti)
Date: Fri, 20 May 2011 02:54:31 -0400
Subject: File AIO and multiple nginx instances
In-Reply-To: <deaa25c9a3b732a5c67d62a2a7338373.NginxMailingListEnglish@forum.nginx.org>
References: <deaa25c9a3b732a5c67d62a2a7338373.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <55e75e5fcfcc8847d08de521525d9d1b.NginxMailingListEnglish@forum.nginx.org>

So, is it a problem of CentOS or Nginx?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198945,199437#msg-199437



From igor at sysoev.ru  Fri May 20 10:56:55 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 20 May 2011 10:56:55 +0400
Subject: File AIO and multiple nginx instances
In-Reply-To: <55e75e5fcfcc8847d08de521525d9d1b.NginxMailingListEnglish@forum.nginx.org>
References: <deaa25c9a3b732a5c67d62a2a7338373.NginxMailingListEnglish@forum.nginx.org>
	<55e75e5fcfcc8847d08de521525d9d1b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110520065655.GD73208@sysoev.ru>

On Fri, May 20, 2011 at 02:54:31AM -0400, arti wrote:
> So, is it a problem of CentOS or Nginx?

I believe this is a CentOS issue, and man page does not allow even
to understand what is the issue.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May 20 10:59:50 2011
From: nginx-forum at nginx.us (arti)
Date: Fri, 20 May 2011 02:59:50 -0400
Subject: File AIO and multiple nginx instances
In-Reply-To: <20110520065655.GD73208@sysoev.ru>
References: <20110520065655.GD73208@sysoev.ru>
Message-ID: <1a2f3404dc4a9235351601fff53a16c7.NginxMailingListEnglish@forum.nginx.org>

Ok, thanks

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198945,199439#msg-199439



From mdounin at mdounin.ru  Fri May 20 12:00:06 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 20 May 2011 12:00:06 +0400
Subject: Disabling basic_auth with rewrites
In-Reply-To: <20110519191656.GA48682@sysoev.ru>
References: <20110519140950.GB42265@mdounin.ru>
	<9739b40539caebc9cbb3d3c9dc3cdecd.NginxMailingListEnglish@forum.nginx.org>
	<20110519181059.GD42265@mdounin.ru>
	<20110519191656.GA48682@sysoev.ru>
Message-ID: <20110520080006.GE42265@mdounin.ru>

Hello!

On Thu, May 19, 2011 at 11:16:56PM +0400, Igor Sysoev wrote:

> On Thu, May 19, 2011 at 10:10:59PM +0400, Maxim Dounin wrote:
> > Hello!
> > 
> > On Thu, May 19, 2011 at 12:43:03PM -0400, klausi wrote:
> > 
> > > Maxim Dounin Wrote:
> > > -------------------------------------------------------
> > > > 
> > > >     location / {
> > > >         auth_basic "protected";
> > > >         auth_basic_user_file
> > > > /etc/nginx/htpasswd/protected;
> > > >         ...
> > > > 
> > > >         location ~ \.php$ {
> > > >             fastcgi_pass ...
> > > >             ...
> > > >         }
> > > >     }
> > > > 
> > > >     location /feeds/importer/ {
> > > >         ...
> > > > 
> > > >         location ~ \.php$ {
> > > >             fastcgi_pass ...
> > > >             ...
> > > >         }
> > > >     }
> > > 
> > > Thanks for the quick reply, nested locations are nice, but they do not
> > > help in this special case. A request to /feeds/importer/* has to be
> > > rewritten to /index.php?q=feeds/importer/* and that should not be
> > > protected. Is unprotecting a path with a special query possible at all?
> > 
> > Ah, sorry, I missed you actually want /feeds/importer/... to be 
> > fully handled by index.php.  This makes configuration even 
> > simplier:
> > 
> >     location / {
> >         auth_basic ...
> >         ...
> > 
> >         location ~ \.php$ {
> >             fastcgi_pass ...
> >             ...
> >         }
> >     }
> > 
> >     location /feeds/importer/ {
> >         rewrite ^/(.*) /index.php?q=$1? break;
> > 
> >         fastcgi_pass ...
> >         ...
> >     }
> >  
> > Note that the only goal of rewrite is to properly change url while 
> > correctly escaping new arguments and stripping old ones (note 
> > trailing '?'), as you probably don't want to allow unauthenticated 
> > users to supply arbitrary arguments to your index.php.  Due to 
> > 'break' request doesn't leave the location in question after 
> > rewrite and processed there.
> 
> My suggestion is to not use rewrite at all:
> 
>      location /feeds/importer/ {
>          location ~ ^/(.*) {
>              fastcgi_pass    ...
>              fastcgi_param   SCRIPT_FILENAME  /path/to/index.php;
>              fastcgi_param   QUERY_STRING     q=$1;
>              ...
>          }

There is a problem: you need urlescape() function then.  Else 
request like

/feeds/importer/&doevil=1

will naturally do evil, i.e. "doevil=1" will be seen by index.php 
as a separate argument.

Maxim Dounin


From mdounin at mdounin.ru  Fri May 20 13:40:33 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 20 May 2011 13:40:33 +0400
Subject: [NGINX] cache process manager consume a lot of CPU at reload and
	has to be restarted
In-Reply-To: <5574016561bfd454ad04bd477e558e06.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTimiy-i0gg1e0nW9wN5rySrLwhkcsA@mail.gmail.com>
	<5574016561bfd454ad04bd477e558e06.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110520094033.GF42265@mdounin.ru>

Hello!

On Thu, May 19, 2011 at 03:51:11PM -0400, omega8cc wrote:

> Unfortunately, it still happens also for 1.0.2 release.
> 
> I have seen it over 10 times on two different servers since upgrade to
> 1.0.0, then 1.0.1 and 1.0.2.
> 
> Not sure how to debug this, as it is random issue, yet causes downtime
> until our auto-healing will kill all nginx processes and wait for load
> to stabilize to start nginx again.

There were two issues discussed in this thread which caused cpu 
hog in cache manager:

1. Earlier SIGSEGV in worker process while holding cache zone lock 
with resulting ngx_spinlock() loop in cache manager.  Fix is to
trace and fix this earlier SIGSEGV (looking though error log 
should help, btw).

2. Cache manager was looping if it wasn't able to made any 
progress.  This was fixed in 1.0.0 (patch present in this thread 
was committed).

I suspect your problem is (1), though it may be some other problem 
as well.  Please provide at least backtrace of runaway process to 
make further debugging possible.  This may be done with

# gdb /path/to/nginx <pid-of-runaway-process>
bt

(and please make sure your binary isn't stripped, else backtrace 
will contain only '??' instead of function names)

Running something like ktrace/strace may be also helpfull.

Maxim Dounin


From mdounin at mdounin.ru  Fri May 20 14:20:56 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 20 May 2011 14:20:56 +0400
Subject: Randomly stopping large file downloads?
In-Reply-To: <afcd4c4d19140b43df79d82ea9980087.NginxMailingListEnglish@forum.nginx.org>
References: <afcd4c4d19140b43df79d82ea9980087.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110520102056.GG42265@mdounin.ru>

Hello!

On Thu, May 19, 2011 at 04:34:34PM -0400, nickn wrote:

> Hi,
> 1.) What would cause nginx to randomly stop serving large file
> downloads? When I try to download a 800MB file it stops serving the file
> after between 10 and 30 MB. I've had to revert to lighttpd because of
> this.

Are you serving file as static one with nginx, or proxy_pass to 
some backend server?  In later case most likely reason is out of 
space (or no access to) in proxy_temp_path.

> I checked the usual things: There wasn't anything in the error log, the
> system had plenty of memory (several hundred MB free) and the worker
> processes were all still there after the download dies. It doesn't
> matter what browser I use.
> 
> Any idea what would cause this?

Please make sure you have error_log properly configured and you 
are looking at right one.  Most simple aproach is to configure one 
at global level and remove others if any (i.e. at http, server, 
etc. levels), e.g.

error_log /path/to/error.log info;

Note that setting logging level to 'info' may be important in some 
cases (e.g. if nginx thinks client timed out or so on).

If the above wouldn't be enough to solve you problem, please 
follow generic instructions as outlined at 
http://wiki.nginx.org/Debugging, i.e. please provide

1. nginx -V output
2. full nginx config
3. debug log

Maxim Dounin


From mdounin at mdounin.ru  Fri May 20 14:23:09 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 20 May 2011 14:23:09 +0400
Subject: try_files
In-Reply-To: <4DD5953C.7030202@styleflare.com>
References: <4DD5953C.7030202@styleflare.com>
Message-ID: <20110520102309.GH42265@mdounin.ru>

Hello!

On Thu, May 19, 2011 at 06:10:04PM -0400, David J. wrote:

> I have a uwsgi_app
> 
> I was wondering why
> 
> location / {
> root /myapp;
> index index.html index.wsgi;
> try_files $uri $uri/ @wsgi;
> }
> ;
> 
> When I visit the domain.com  returns a 403 (when I dont have an
> index.html file in the root directory);
> 
> but if I do domain.com/index.wsgi
> 
> The WSGI app is called. This all works.
> 
> I am guessing I am doing something wrong.
> 
> In my config I have
> 
> location @wsgi {
>     include uwsgi_params;
>     uwsgi_pass unix:/tmp/uwsgi.sock
> }

Please show full config.

Maxim Dounin


From david at styleflare.com  Fri May 20 14:57:23 2011
From: david at styleflare.com (David J.)
Date: Fri, 20 May 2011 06:57:23 -0400
Subject: try_files
In-Reply-To: <20110520102309.GH42265@mdounin.ru>
References: <4DD5953C.7030202@styleflare.com>
	<20110520102309.GH42265@mdounin.ru>
Message-ID: <4DD64913.6080509@styleflare.com>

Running Nginx v1.0.0

Here is server config

server {
         listen       80;
         server_name domain.com www.domain.com;
location /{
index index.html index index.htm index.php;
root   /www/domain.com;
try_files $uri $uri/ @wsgi;
}

location @wsgi {
index  index.html index.htm;
include uwsgi_params;

uwsgi_pass 
unix:/tmp/uwsgi.sock;                                                                                                                                                               
}
                                                                                                                                                                         location  ~ \.css$ {
root   /www/domain.com;
}

}


On 5/20/11 6:23 AM, Maxim Dounin wrote:
> Hello!
>
> On Thu, May 19, 2011 at 06:10:04PM -0400, David J. wrote:
>
>> I have a uwsgi_app
>>
>> I was wondering why
>>
>> location / {
>> root /myapp;
>> index index.html index.wsgi;
>> try_files $uri $uri/ @wsgi;
>> }
>> ;
>>
>> When I visit the domain.com  returns a 403 (when I dont have an
>> index.html file in the root directory);
>>
>> but if I do domain.com/index.wsgi
>>
>> The WSGI app is called. This all works.
>>
>> I am guessing I am doing something wrong.
>>
>> In my config I have
>>
>> location @wsgi {
>>      include uwsgi_params;
>>      uwsgi_pass unix:/tmp/uwsgi.sock
>> }
> Please show full config.
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>



From igor at sysoev.ru  Fri May 20 15:06:58 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 20 May 2011 15:06:58 +0400
Subject: Proxy_pass and slash witch nginx
In-Reply-To: <0cbde934384bd99d3e3d381c2346af82.NginxMailingListEnglish@forum.nginx.org>
References: <20110518195045.GX42265@mdounin.ru>
	<0cbde934384bd99d3e3d381c2346af82.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110520110658.GA95649@sysoev.ru>

On Thu, May 19, 2011 at 04:37:38PM -0400, myckeul wrote:
> Hello,
> 
> Thanks for your response but they won't working.

> my vhost : 
> server {
>        listen [::]:80;
> 
>        server_name proxy.com *.proxy.com;
> 
>        access_log  /var/log/nginx/proxy_.access.log;
> 
>        error_page 404 = /404.htm;
> 
>        location /404.htm {
>        	root /var/www;
>        }
> 
>        location / {
>        		proxy_set_header X-Forwarded-Host $host;
> 		proxy_set_header X-Forwarded-Server $host;
> 		proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
> 
> 
> 		if ($host ~* ^proxy\.com$) {
> 		   proxy_pass http://mydomain.com;
> 		}
> 
> 		if ($host ~* ^www\.proxy\.com$) {
> 		   proxy_pass http://www.mydomain.com;
> 		}
> 
> 		if ($host ~* ^subdomain\.proxy\.com$) {
> 		   proxy_pass http://subdomain.mydomain.com;
> 		}
> 
> 		proxy_redirect http://proxy.com/test http://proxy.com/test/;
> 	}
> }

proxy_redirect http://mydomain.com http://proxy.com;
proxy_redirect http://www.mydomain.com http://www.proxy.com;
proxy_redirect http://subdomain.mydomain.com http://subdomain.proxy.com;

BTW, your configuraiton is wrong mix of levels: you should not use
virtual host name on location level in "if ($host ~". This leads
to cumbersome and unmaintainable configurations.
Instead you should define 3 separate servers:

server {
      server_name  proxy.com;
      location / {
          proxy_pass  http://mydomain.com;
      }
      ...
}

server {
      server_name  www.proxy.com;
      location / {
          proxy_pass  http://www.mydomain.com;
      }
      ...
}

server {
      server_name  subdomain.proxy.com;
      location / {
          proxy_pass  http://subdomain.mydomain.com;
      }
      ...
}


-- 
Igor Sysoev


From igor at sysoev.ru  Fri May 20 15:19:52 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 20 May 2011 15:19:52 +0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <e00b353ec548b2c91b169c5b8ccaaa9c.NginxMailingListEnglish@forum.nginx.org>
References: <c0c6df7135604bcee44a098e5b896a30@ruby-forum.com>
	<e00b353ec548b2c91b169c5b8ccaaa9c.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110520111952.GC95649@sysoev.ru>

On Wed, May 18, 2011 at 10:52:39PM -0400, ltss wrote:
> Hello i am new to nginx and i need some help to convert my .htaccess, i
> have tried some shit but that not work ...
> 
> 	Options FollowSymLinks
> 	RewriteEngine On
> 
> 	RewriteRule ^(.*)\.php/img/(.*)$ img/$2 [C]
> 	RewriteRule ^/img/(.*)$ img/$2 [QSA,L]
> 
> 	RewriteCond %{REQUEST_FILENAME} !^scriptsFtp.*$
> 
> 	RewriteCond %{REQUEST_FILENAME} !-d
> 	RewriteCond %{REQUEST_FILENAME} !-f
> 		RewriteRule ^(.*)$ index.php [QSA,L]

It's not easy to understand what these .htaccess should do.
Probably this:

root  /path/to/files;

location / {
   try_files  $uri  $uri/  /index.php?$args;
}

location /scriptsFtp {
   ...
}

location ~ ^(.*)\.php/img/(.*)$ {
   alias  /path/to/img/$2;
}

location ~ .php$ {
   fastcgi_pass   ...
   ...
}


-- 
Igor Sysoev


From Richard.Kearsley at m247.com  Fri May 20 15:32:56 2011
From: Richard.Kearsley at m247.com (Richard Kearsley)
Date: Fri, 20 May 2011 11:32:56 +0000
Subject: lua location capture
In-Reply-To: <BANLkTi=_eN=EzKF-J9=cFgu6qd29GVOwuA@mail.gmail.com>
References: <EFC5A19DBF3A2E489904733B467D990B12493E@office-server2.m247.local>
	<BANLkTi=_eN=EzKF-J9=cFgu6qd29GVOwuA@mail.gmail.com>
Message-ID: <EFC5A19DBF3A2E489904733B467D990B125375@office-server2.m247.local>

Hi
Thanks for the reply
I traced the issue with the header's not being re-sent

ngx.location.capture("/some-location", { method = ngx.HTTP_HEAD })
they are not re-sent

but
ngx.location.capture("/some-location", { method = ngx.HTTP_GET })
they are sent

I don't know if this is correct behavior (it may well be)
I got around it by changing the headers I needed into args then re-creating them at /some-location :(

-----Original Message-----
From: agentzh [mailto:agentzh at gmail.com] 
Sent: 20 May 2011 04:56
To: nginx at nginx.org
Subject: Re: lua location capture

On Fri, May 20, 2011 at 5:59 AM, Richard Kearsley
<Richard.Kearsley at m247.com> wrote:
> Hi
>
> The lua doc at https://github.com/chaoslawful/lua-nginx-module
>
> Says "Note that, by default, subrequests issued by ngx.location.capture
> inherit all the request headers of the current request."
>
> And goes on to say that it can be an unwanted feature
>
> In my project I actually take advantage of this feature, but for some reason
> the headers don't go through to the sub request
>

Could you please paste your nginx.conf here? Also, information about
your nginx's version and ngx_lua's version can also be helpful :)

Best,
-agentzh

_______________________________________________
nginx mailing list
nginx at nginx.org
http://nginx.org/mailman/listinfo/nginx


From mdounin at mdounin.ru  Fri May 20 15:33:15 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 20 May 2011 15:33:15 +0400
Subject: 2 questions on send_timeout (maybe bugs)
In-Reply-To: <76e2ab59.13e9e.1300b8dcee7.Coremail.wubingzheng@163.com>
References: <20110519113842.GA42265@mdounin.ru>
	<20110518133507.GV42265@mdounin.ru>
	<10a6803e.1907c.130026300cb.Coremail.wubingzheng@163.com>
	<54e4f427.3d5f.130061f7663.Coremail.wubingzheng@163.com>
	<76e2ab59.13e9e.1300b8dcee7.Coremail.wubingzheng@163.com>
Message-ID: <20110520113314.GI42265@mdounin.ru>

Hello!

On Fri, May 20, 2011 at 11:58:20AM +0800, Wu Bingzheng wrote:

> Hi Maxim,
> 
> 1.
> I am so sorry that I made a big mistake on question 1.
> Nginx didn't detect the downstream breaking, because I use "proxy_ignore_client_abort on;".
> So Nginx works correctlly. I am so sorry about it.

I believe the problem may still appear with normal usage once 
upstream connection terminates, EPOLLHUP handling in epoll event 
module doesn't look correct for me.  This should be investigated 
further.

> 2.
> For question 2.
> You said that :
> >Note also that in debug log you provided nginx correctly times out 
> >after 20s from last send.
> 
> That's because I set the send_timeout as 10s. I am sorry that I did not make clear of that.
> 
> You can reproduce the doubled timeout easy:

Ok, I was able to reproduce this and see the problem.  Indeed, 
when upstream module detects client timeout it doesn't finalize 
request properly, and as a result http core thinks it should 
continue sending data in buffers and hence waits for another 
timeout.

Quick fix for this particular problem would be something like

diff --git a/src/http/ngx_http_upstream.c b/src/http/ngx_http_upstream.c
--- a/src/http/ngx_http_upstream.c
+++ b/src/http/ngx_http_upstream.c
@@ -2320,7 +2320,7 @@ ngx_http_upstream_process_non_buffered_d
     if (wev->timedout) {
         c->timedout = 1;
         ngx_connection_error(c, NGX_ETIMEDOUT, "client timed out");
-        ngx_http_upstream_finalize_request(r, u, 0);
+        ngx_http_upstream_finalize_request(r, u, NGX_HTTP_REQUEST_TIME_OUT);
         return;
     }

@@ -3013,6 +3013,7 @@ ngx_http_upstream_finalize_request(ngx_h
 #endif

     if (u->header_sent
+        && rc != NGX_HTTP_REQUEST_TIME_OUT
         && (rc == NGX_ERROR || rc >= NGX_HTTP_SPECIAL_RESPONSE))
     {
         rc = 0;


Though upstream module certainly requires larger cleanup of error 
handling.  This is not the first time I've been bitten by upstream 
not finalizing requests properly.

Maxim Dounin

> 
> [server] using nginx-1.0.0, without any addon module. in nginx.conf:
> 
>     server {
>         listen       5821;
>         proxy_buffering off; #must in non-buffering mode
>         send_timeout 10;
> 
>         location / {
>             proxy_pass http://sourceserver;
>         }
>     }
> 
> 
> 
> [client] you can run this simple python script:
> 
>   #!/usr/bin/python
>   import socket
>   import time
>   s = socket.socket(socket.AF_INET,socket.SOCK_STREAM)
>   s.connect((127.0.0.1, 5821))
>   s.send("GET /data/something HTTP/1.0\r\n\r\n")
>   buf = s.recv(100)
>   print buf
>   time.sleep(100)
> 
> 
> 
> Then, in access.log, you will find that the request is finished in 20s.
> 
> 
> Best Regads,
> Wu
> 
> 
> 
> 
> 
> 
> 
> 
> At 2011-05-19 19:38:42?"Maxim Dounin" <mdounin at mdounin.ru> wrote:
> 
> >Hello!
> >
> >On Thu, May 19, 2011 at 10:39:43AM +0800, Wu Bingzheng wrote:
> >
> >> Hi Maxim,
> >> 
> >> Thanks for your answer. But I am still confused.
> >> 
> >> 
> >> At 2011-05-18 21:35:07?"Maxim Dounin" <mdounin at mdounin.ru> wrote:
> >> 
> >> >Hello!
> >> >
> >> >On Wed, May 18, 2011 at 05:15:00PM +0800, Wu Bingzheng wrote:
> >> >
> >> >> hi all,
> >> >> 
> >> >> 
> >> >> I meet 2 question when using nginx 1.0.0.
> >> >> 
> >> >> 
> >> >> Question1.
> >> >> 
> >> >> what happens if the downstream connection breaks, before all 
> >> >> data sent to the client?
> >> >> 
> >> >> 1. get client's request
> >> >> 2. set handler: r->read_event_handle=ngx_http_block_reading
> >> >> 3. send data to client, and blocked ( 
> >> >> write_filter.c:c->send_chain() returns non-NULL or non-error), 
> >> >> and go back to epoll_wait
> >> >> 4. HERE, the downstream breaks
> >> >> 5. epoll detect the breaking, so epoll_wait() continues in 
> >> >> ngx_epoll_process_events(), where the revents==EPOLL_IN | 
> >> >> EPOLL_ERR | EPOLL_HUP. So we call rev->handler(), which calls 
> >> >> r->read_event_handle, which dose nothing. AS A RESULT, the 
> >> >> downstream breaking was not detected by nginx !!!
> >> >
> >> >Here write event handler will be called.  While standard 
> >> >ngx_http_writer() is in use - it will call output filter chain and 
> >> >detect error.  Else it's up to active handler to correctly handle 
> >> >this, e.g. standard upstream modules (proxy/fastcgi/...) may 
> >> >either terminate request immediately or wait for full upstream 
> >> >response depending on various settings (proxy_ignore_client_abort, 
> >> >proxy_cache/proxy_store enabled).
> >> >
> >> 
> >> You said the write handler will be called. But if a connection 
> >> is break, revents returned by epoll_wait() is 'EPOLL_IN |  
> >> EPOLL_ERR | EPOLL_HUP', and no EPOLL_OUT in revents. So I think 
> >> the write-handler will NOT be called.
> >
> >Ah, ok, it looks like I was wrong here.  Write handler will be 
> >only called if there are no EPOLLIN flag in revents, and this is 
> >probably wrong assuming EPOLLHUP follows the same semantics as 
> >POLLHUP[1], i.e.
> >
> >POLLHUP
> >...  This event and POLLOUT are mutually-exclusive; a stream can 
> >never be writable if a hangup has occurred. ...
> >
> >Could you please provide detailed instructions to reproduce the 
> >problem, preferably with standard modules involved?
> >
> >[1] http://pubs.opengroup.org/onlinepubs/9699919799/functions/poll.html
> >
> >> The FACT is, in my server, there are many many access logs whose 
> >> time-taken are 40.001s. (send_timeout is 20s, and this is a 
> >> music online server, so many people re-choose a song before the 
> >> current song finishs, which means the downstream breaks before 
> >> nginx sends all data).
> >> 
> >> I add 2 debug log in ngx_epoll_process_events(), and copy part 
> >> of the debug log here.
> >
> >Just a side note: mostly same information may be seen at global 
> >level debug log.
> >
> >[...]
> >
> >> >> I set 'send_timeout' as 20s, but nginx timeouts in 40s actually.
> >> >> 
> >> >> 1. in 
> >> >> src/http/ngx_http_upstream.c[line2448]:ngx_http_upstream_process_non_buffered_request(), 
> >> >> a timer is added;
> >> >> 2. after that timer timeout, in 
> >> >> src/http/ngx_http_request.c[line2179]:ngx_http_set_write_handler(), 
> >> >> another timer is added again.
> >> >> 
> >> >> so the send_timeout is doubled.
> >> >
> >> >Timeout is re-set on each send operation, and expected to be 
> >> >triggered when nothing happens during send_timeout.
> >> 
> >> So, you think that, I set 'send_timeout' as 20s, but nginx 
> >> actually timeout in 40s is right and expected?
> >
> >It should timeout in 20s after last send.  Whether request time 
> >of 40s is wrong or not depends on many factors, but the code you 
> >pointed out looks fine.
> >
> >Note also that in debug log you provided nginx correctly times out 
> >after 20s from last send.
> >
> >Maxim Dounin
> 
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx


From mdounin at mdounin.ru  Fri May 20 16:12:27 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 20 May 2011 16:12:27 +0400
Subject: try_files
In-Reply-To: <4DD64913.6080509@styleflare.com>
References: <4DD5953C.7030202@styleflare.com>
	<20110520102309.GH42265@mdounin.ru>
	<4DD64913.6080509@styleflare.com>
Message-ID: <20110520121227.GJ42265@mdounin.ru>

Hello!

On Fri, May 20, 2011 at 06:57:23AM -0400, David J. wrote:

> Running Nginx v1.0.0
> 
> Here is server config
> 
> server {
>         listen       80;
>         server_name domain.com www.domain.com;
> location /{
> index index.html index index.htm index.php;
> root   /www/domain.com;
> try_files $uri $uri/ @wsgi;
> }
> 
> location @wsgi {
> index  index.html index.htm;
> include uwsgi_params;
> 
> uwsgi_pass unix:/tmp/uwsgi.sock;
> }
>                                                                                                                                                                         location  ~ \.css$ {
> root   /www/domain.com;
> }
> 
> }

It doesn't looks like you have "index" directive set to 
"index.wsgi" nor any actual handling of index.wsgi (except 
try_files fallback).  This all makes me think that you actually 
don't have index.wsgi on filesystem either.

Note that "index" directive will actually test existence of given 
index files on filesystem and will do an internal redirect if one 
was found.  That is, if you don't have any index files on 
filesystem - you'll get 403 (unless autoindex module enabled).

The configuration you probably want is something like

    root /www/domain.com;

    location / {
        try_files $uri $uri/index.html @wsgi;
    }

    location @wsgi {
        uwsgi_pass ...
    }

That is: if file exists, return it; if there is a directory with 
index.html, return it; else fallback to wsgi.

Maxim Dounin

> 
> 
> On 5/20/11 6:23 AM, Maxim Dounin wrote:
> >Hello!
> >
> >On Thu, May 19, 2011 at 06:10:04PM -0400, David J. wrote:
> >
> >>I have a uwsgi_app
> >>
> >>I was wondering why
> >>
> >>location / {
> >>root /myapp;
> >>index index.html index.wsgi;
> >>try_files $uri $uri/ @wsgi;
> >>}
> >>;
> >>
> >>When I visit the domain.com  returns a 403 (when I dont have an
> >>index.html file in the root directory);
> >>
> >>but if I do domain.com/index.wsgi
> >>
> >>The WSGI app is called. This all works.
> >>
> >>I am guessing I am doing something wrong.
> >>
> >>In my config I have
> >>
> >>location @wsgi {
> >>     include uwsgi_params;
> >>     uwsgi_pass unix:/tmp/uwsgi.sock
> >>}
> >Please show full config.
> >
> >Maxim Dounin
> >
> >_______________________________________________
> >nginx mailing list
> >nginx at nginx.org
> >http://nginx.org/mailman/listinfo/nginx
> >
> 
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx


From lists at ruby-forum.com  Fri May 20 21:14:44 2011
From: lists at ruby-forum.com (Yanxin Z.)
Date: Fri, 20 May 2011 19:14:44 +0200
Subject: nginx for multiple sub-domain?
Message-ID: <7136024a70c9eb0cac6f94538240d973@ruby-forum.com>

Hello,
I have a web site, e.g.,  www.example.com
under it, i have several sub-domains, e.g., /register, /query, /report.

I have a requirement that,
/query is for port 80
/register is for port 443, but with server side SSL certificate
verification.
/report is for port 443 , with both server and client side SSL
certificate verification.

I want to know whether NGX config can do this ?


I have the problem to config /register and /report under port 443.

Thanks,
Yanxin

-- 
Posted via http://www.ruby-forum.com/.


From igor at sysoev.ru  Fri May 20 22:00:06 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 20 May 2011 22:00:06 +0400
Subject: nginx for multiple sub-domain?
In-Reply-To: <7136024a70c9eb0cac6f94538240d973@ruby-forum.com>
References: <7136024a70c9eb0cac6f94538240d973@ruby-forum.com>
Message-ID: <20110520180006.GB9846@sysoev.ru>

On Fri, May 20, 2011 at 07:14:44PM +0200, Yanxin Z. wrote:
> Hello,
> I have a web site, e.g.,  www.example.com
> under it, i have several sub-domains, e.g., /register, /query, /report.
> 
> I have a requirement that,
> /query is for port 80
> /register is for port 443, but with server side SSL certificate
> verification.
> /report is for port 443 , with both server and client side SSL
> certificate verification.
> 
> I want to know whether NGX config can do this ?
> 
> 
> I have the problem to config /register and /report under port 443.

server {
    listen  80;
    server_name  www.example.com;
    location /query {
        ...
    }
}

server {
    listen  443;
    server_name  www.example.com;

    ssl                  on;
    ssl_certificate      www.nginx.com.crt;
    ssl_certificate_key  www.nginx.com.key;

    ssl_verify_client       optional;
    ssl_client_certificate  CA.crt;

    location /register {
         ...
    }

    location /report {
         if ($ssl_client_verify != SUCCESS) {
             return 403;
         }
         ...
    }
}

http://nginx.org/en/docs/http/configuring_https_servers.html


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May 20 22:40:10 2011
From: nginx-forum at nginx.us (TECK)
Date: Fri, 20 May 2011 14:40:10 -0400
Subject: Rewrite or internal redirection cycle?
In-Reply-To: <20110520053436.GB73208@sysoev.ru>
References: <20110520053436.GB73208@sysoev.ru>
Message-ID: <0e8e32f8c2b80c2ad13a35f3005a9abc.NginxMailingListEnglish@forum.nginx.org>

Thank you Igor, I will use the SCRIPT_FILENAME setup.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199260,199596#msg-199596



From lists at ruby-forum.com  Fri May 20 22:50:09 2011
From: lists at ruby-forum.com (Yanxin Z.)
Date: Fri, 20 May 2011 20:50:09 +0200
Subject: nginx for multiple sub-domain?
In-Reply-To: <7136024a70c9eb0cac6f94538240d973@ruby-forum.com>
References: <7136024a70c9eb0cac6f94538240d973@ruby-forum.com>
Message-ID: <028c9c62aa3970a88144d3fbfb996d2f@ruby-forum.com>

Thank you Igor!!
NGX is so powerful. It's working for me.


Another question, for CA cert, I also need to verify PEM pass phase.
Can I do it in NGX config?

Thanks,
Yanxin

-- 
Posted via http://www.ruby-forum.com/.


From nginx-forum at nginx.us  Fri May 20 23:43:53 2011
From: nginx-forum at nginx.us (nickn)
Date: Fri, 20 May 2011 15:43:53 -0400
Subject: Randomly stopping large file downloads?
In-Reply-To: <20110520102056.GG42265@mdounin.ru>
References: <20110520102056.GG42265@mdounin.ru>
Message-ID: <908b73582b85537775efe9c692b948c5.NginxMailingListEnglish@forum.nginx.org>

Hi Maxim, thanks for your help.
I'm not using a proxy, it's just a straight connection to nginx.

I ran it again with error log on info and found that it thinks the
client is timing out, although watching firefox it seemed to be
downloading ok the entire time and I wasn't having connectivity issues
on other sites. 

I tried keepalive_timeout  65; rather than 5 but had the same problem.

Here's the info below, any ideas what it could be?

----

/var/log/nginx# tail cosmic-access.log
76.250.143.130 - - [20/May/2011:19:32:50 +0000] "GET /bigfile.sql.gz
HTTP/1.1" 200 114499284 "-" "Mozilla/5.0 (Windows; U; Windows NT 5.1;
en-US; rv:1.9.2.17) Gecko/20110420 Firefox/3.6.17"

/var/log/nginx# tail cosmic-error.log
2011/05/20 19:32:50 [info] 12222#0: *10 client timed out (110:
Connection timed out) while sending response to client, client:
76.250.143.130, server: cosmicchannelings.com, request: "GET
/bigfile.sql.gz HTTP/1.1", host: "cosmicchannelings.com:81"


root at 217948:/var/log/nginx# nginx -V
nginx version: nginx/0.7.65
TLS SNI support enabled
configure arguments: --conf-path=/etc/nginx/nginx.conf
--error-log-path=/var/log/nginx/error.log --pid-path=/var/run/nginx.pid
--lock-path=/var/lock/nginx.lock
--http-log-path=/var/log/nginx/access.log
--http-client-body-temp-path=/var/lib/nginx/body
--http-proxy-temp-path=/var/lib/nginx/proxy
--http-fastcgi-temp-path=/var/lib/nginx/fastcgi --with-debug
--with-http_stub_status_module --with-http_flv_module
--with-http_ssl_module --with-http_dav_module
--with-http_gzip_static_module --with-http_realip_module --with-mail
--with-mail_ssl_module --with-ipv6
--add-module=/build/buildd/nginx-0.7.65/modules/nginx-upstream-fair


---------
/etc/nginx/nginx.conf:
---------

user www-data;
#worker_processes  1;
worker_processes  2;

pid        /var/run/nginx.pid;

events {
    worker_connections  1024;
    # multi_accept on;
}

http {
    include       /etc/nginx/mime.types;

    error_log  /var/log/nginx/error.log info;
    access_log	/var/log/nginx/access.log;

    sendfile        on;
    #tcp_nopush     on;

    #keepalive_timeout  0;
    #keepalive_timeout  65;
    keepalive_timeout  5;
    tcp_nodelay        on;

#    gzip  on;
#    gzip_disable "MSIE [1-6]\.(?!.*SV1)";

    include /etc/nginx/conf.d/*.conf;
    include /etc/nginx/sites-enabled/*;
}

# mail {
#     # See sample authentication script at:
#     # http://wiki.nginx.org/NginxImapAuthenticateWithApachePhpScript
# 
#     # auth_http localhost/auth.php;
#     # pop3_capabilities "TOP" "USER";
#     # imap_capabilities "IMAP4rev1" "UIDPLUS";
# 
#     server {
#         listen     localhost:110;
#         protocol   pop3;
#         proxy      on;
#     }
# 
#     server {
#         listen     localhost:143;
#         protocol   imap;
#         proxy      on;
#     }
# }


---------------
/etc/nginx/sites-enabled/sites-enabled/cosmic:
---------------


# You may add here your
# server {
#	...
# }
# statements for each of your virtual hosts

server {
	listen   81;
	server_name  cosmicchannelings.com www.cosmicchannelings.com;

#	access_log  /var/log/apache2/cosmic-access.log;
#	error_log  /var/log/apache2/cosmic-error.log;
	access_log  /var/log/nginx/cosmic-access.log;
	error_log  /var/log/nginx/cosmic-error.log info;

	root   /var/www/cosmic;
	index  index.php index.html index.htm;

        # wordpress url writing
        error_page  404              /blog/index.php;

	if ($http_user_agent ~ "magpie-crawler" ) { return 403; }
 
        # prevents nginx from returning 404s all the time along with
good content
        try_files $uri $uri/ /blog/index.php;

        # also works, may be slower?
        # if (!-e $request_filename) {
        #        rewrite ^/blog/(.+)$ /blog/index.php?q=$1 last;
        # }

# cave yahoo group
#    "^/i/([a-z0-9_\-]+)/([a-z0-9_\-]+)/([a-z0-9_\-]+)?$" =>
"/cave/link.php?m=$1&g=$2&s=$3",
rewrite ^/i/([a-z0-9_\-]+)/([a-z0-9_\-]+)/([a-z0-9_\-]+)?$
/cave/link.php?m=$1&g=$2&s=$3 permanent;

# convert old blog url format to new. old had post id, new only has
slug
#    "^(/blog/[a-z0-9_\-]+)/[0-9]+$" => "$1",
rewrite ^(/blog/[a-z0-9_\-]+)/[0-9]+$ $1 permanent;

# fix broken link errors. /blog/k_karma_29/family.html =>
/blog/k_family
#    "^/blog/k_[a-z0-9_\-]+/([a-z0-9_\-]+)\.html+$" => "/blog/k_$1",
rewrite ^/blog/k_[a-z0-9_\-]+/([a-z0-9_\-]+)\.html+$ /blog/k_$1
permanent;

# /blog/tag/new-world-order/topic.html
#    "^/blog/tag/[a-z0-9_\-]+/([a-z0-9_\-]+)\.html+$" => "/blog/k_$1",
rewrite ^/blog/tag/[a-z0-9_\-]+/([a-z0-9_\-]+)\.html+$ /blog/k_$1
permanent;

# redirect old archive html pages to blog pages
#    "^/archive/topic/([a-z0-9_\-]+)\.html+$" => "/blog/k_$1",
rewrite ^/archive/topic/([a-z0-9_\-]+)\.html+$ /blog/k_$1 permanent;

# fix bug where topic lines in keyword pages got linked as
http://www.cosmicchannelings.com/blog/heaven.html
#    "^/blog/([a-z0-9_\-]+)\.html+$" => "/blog/k_$1"
rewrite ^/blog/([a-z0-9_\-]+)\.html+$ /blog/k_$1 permanent;
 

	# redirect server error pages to the static page /50x.html
	#
	#error_page   500 502 503 504  /50x.html;
	#location = /50x.html {
	#	root   /var/www/nginx-default;
	#}

	# proxy the PHP scripts to Apache listening on 127.0.0.1:80
	#
	#location ~ \.php$ {
		#proxy_pass   http://127.0.0.1;
	#}

	# pass the PHP scripts to FastCGI server listening on 127.0.0.1:9000
	#
    location ~ .php$ {
        fastcgi_split_path_info ^(.+\.php)(.*)$;
        fastcgi_pass   backend;
        fastcgi_index  index.php;
        #fastcgi_param  SCRIPT_FILENAME 
/var/www/cosmic$fastcgi_script_name;
        fastcgi_param  SCRIPT_FILENAME
$document_root$fastcgi_script_name;

        include fastcgi_params;
        fastcgi_param  QUERY_STRING     $query_string;
        fastcgi_param  REQUEST_METHOD   $request_method;
        fastcgi_param  CONTENT_TYPE     $content_type;
        fastcgi_param  CONTENT_LENGTH   $content_length;
        fastcgi_intercept_errors        on;
        fastcgi_ignore_client_abort     off;
        fastcgi_connect_timeout 60;
        fastcgi_send_timeout 180;
        fastcgi_read_timeout 180;

        #fastcgi_buffer_size 128k;
        #fastcgi_buffers 4 256k;
	
	# from http://forum.nginx.org/read.php?3,70566
	fastcgi_buffers 256 16k;
	fastcgi_buffer_size 32k;

        fastcgi_busy_buffers_size 256k;
        fastcgi_temp_file_write_size 256k;
    }

	# deny access to .htaccess files, if Apache's document root
	# concurs with nginx's one
	#
	location ~ /\.ht {
		deny  all;
	}
}

upstream backend {
        server 127.0.0.1:9000;
}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199359,199603#msg-199603



From nginx-forum at nginx.us  Sat May 21 00:02:07 2011
From: nginx-forum at nginx.us (ltss)
Date: Fri, 20 May 2011 16:02:07 -0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <e00b353ec548b2c91b169c5b8ccaaa9c.NginxMailingListEnglish@forum.nginx.org>
References: <c0c6df7135604bcee44a098e5b896a30@ruby-forum.com>
	<e00b353ec548b2c91b169c5b8ccaaa9c.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <6856e4d76e54e6ffbc66eaa7c05442ce.NginxMailingListEnglish@forum.nginx.org>

thx i gonna to try it tommorow ;)

the system is used for subdomain with a wildcard, example
toto.domain.com  show a mini website for toto and same things for 300
other users, the website is the same but the content depend of the name
of the subdomain, all is stored in mysql for images and such ...

sorry for my english ...

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,23868,199607#msg-199607



From mdounin at mdounin.ru  Sat May 21 12:51:17 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sat, 21 May 2011 12:51:17 +0400
Subject: Randomly stopping large file downloads?
In-Reply-To: <908b73582b85537775efe9c692b948c5.NginxMailingListEnglish@forum.nginx.org>
References: <20110520102056.GG42265@mdounin.ru>
	<908b73582b85537775efe9c692b948c5.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110521085116.GM42265@mdounin.ru>

Hello!

On Fri, May 20, 2011 at 03:43:53PM -0400, nickn wrote:

> Hi Maxim, thanks for your help.
> I'm not using a proxy, it's just a straight connection to nginx.
> 
> I ran it again with error log on info and found that it thinks the
> client is timing out, although watching firefox it seemed to be
> downloading ok the entire time and I wasn't having connectivity issues
> on other sites. 
> 
> I tried keepalive_timeout  65; rather than 5 but had the same problem.
> 
> Here's the info below, any ideas what it could be?

Timeout to tune is send_timeout, not keepalive one.  Anyway, 
please provide debug log and tcpdump of the connection in 
question.

[...]

> root at 217948:/var/log/nginx# nginx -V
> nginx version: nginx/0.7.65

Note that 0.7.65 is rather old.  You may want to upgrade to 1.0.2.

[...]

>     sendfile        on;

When using sendfile with big files it's good idea to use something 
like

      sendfile_max_chunk 1m;

as well.

[...]

>     include /etc/nginx/conf.d/*.conf;
>     include /etc/nginx/sites-enabled/*;

If there are any files here except the one you already showed - 
you may want to show them as well (or at least parts which may 
affect server{} in question).

Maxim Dounin


From nginx-forum at nginx.us  Sat May 21 18:09:17 2011
From: nginx-forum at nginx.us (myckeul)
Date: Sat, 21 May 2011 10:09:17 -0400
Subject: Proxy_pass and slash witch nginx
In-Reply-To: <20110520110658.GA95649@sysoev.ru>
References: <20110520110658.GA95649@sysoev.ru>
Message-ID: <7cc085f85985df209205af97d16729d4.NginxMailingListEnglish@forum.nginx.org>

hello ;)

thank you, everything is ok now 

have a nice day

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,198964,199488#msg-199488



From adam at lionite.com  Sun May 22 04:41:59 2011
From: adam at lionite.com (Adam Benayoun)
Date: Sun, 22 May 2011 03:41:59 +0300
Subject: simple rewrite rules
Message-ID: <BANLkTi=gjCwvH6UkYdZawm2JvZbNXiWQdg@mail.gmail.com>

Hello everyone,
I'm trying to setup some basic rewrite rules in nginx (I'm sorry if I may be
asking obvious question but I moved my servers from apache to nginx and I
can't figure out how to write simple rewrite rules in nginx).

This is what I am trying to achieve:

I have a URL with the following arguments:

http://www.domain.com/browse/category

What I want to achieve is:
- When someone try to reach http://www.domain.com/category - it will serve
the url http://www.domain.com/browse/category internally (no 301 redirect).
- When someone will try to reach http://www.domain.com/browse/category it
will redirect externally (301) to http://www.domain.com/category

I've added 2 rewriterule at first but somehow only the first rule works, the
second won't work:

        rewrite ^/browse/(.*) http://www.domain.com/$1 permanent;
        rewrite ^/(category1|category2|category3|category4)$    /browse/$1
last;

Thanks (and sorry for asking a somewhat newbie question!)

p.s: yes, I've read this http://wiki.nginx.org/HttpRewriteModule
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110522/b54638dc/attachment.html>

From nginx-forum at nginx.us  Sun May 22 08:35:07 2011
From: nginx-forum at nginx.us (yaz)
Date: Sun, 22 May 2011 00:35:07 -0400
Subject: My first rewrite QUERY_STRING
Message-ID: <c50c4c6d2e5653978c341e8a3dc62427.NginxMailingListEnglish@forum.nginx.org>

Hello 

here is a piece of htaccess i am stuck converting into nginx rewrite
conventions : 

RewriteCond %{QUERY_STRING} ^sms_ss= [OR]
RewriteCond %{QUERY_STRING} ^scvds=
RewriteRule ^(.*)$ http://www.mysite.com/$1? [R=301,L]


I am trying something like 

 if ($args ~ "sms_ss= (.*)") {
        rewrite ^ /(.*)=$args? break;
    }

Not working 


Any advice would be appreciated thanks

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199820,199820#msg-199820



From igor at sysoev.ru  Sun May 22 09:21:56 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Sun, 22 May 2011 09:21:56 +0400
Subject: simple rewrite rules
In-Reply-To: <BANLkTi=gjCwvH6UkYdZawm2JvZbNXiWQdg@mail.gmail.com>
References: <BANLkTi=gjCwvH6UkYdZawm2JvZbNXiWQdg@mail.gmail.com>
Message-ID: <20110522052155.GB65681@sysoev.ru>

On Sun, May 22, 2011 at 03:41:59AM +0300, Adam Benayoun wrote:
> Hello everyone,
> I'm trying to setup some basic rewrite rules in nginx (I'm sorry if I may be
> asking obvious question but I moved my servers from apache to nginx and I
> can't figure out how to write simple rewrite rules in nginx).
> 
> This is what I am trying to achieve:
> 
> I have a URL with the following arguments:
> 
> http://www.domain.com/browse/category
> 
> What I want to achieve is:
> - When someone try to reach http://www.domain.com/category - it will serve
> the url http://www.domain.com/browse/category internally (no 301 redirect).
> - When someone will try to reach http://www.domain.com/browse/category it
> will redirect externally (301) to http://www.domain.com/category
> 
> I've added 2 rewriterule at first but somehow only the first rule works, the
> second won't work:
> 
>         rewrite ^/browse/(.*) http://www.domain.com/$1 permanent;
>         rewrite ^/(category1|category2|category3|category4)$    /browse/$1
> last;
> 
> Thanks (and sorry for asking a somewhat newbie question!)
> 
> p.s: yes, I've read this http://wiki.nginx.org/HttpRewriteModule

You should not read anyting about rewrite module. It's better to forget
about it. You should think using locations.

     location /category1/
         #depending on processing method:

         # FastCGI
         fastcgi_pass    backend;
         fatscgi_param   SCRIPT_FILENMAE   /path/to/browse;

         # proxying
         proxy_pass    http://backend/browse/

         # static files
         root   /path/to/browse;
     }

     location /browse/ {
         rewrite ^/browse/(.*)  /$1  permanent;
     }


-- 
Igor Sysoev


From igor at sysoev.ru  Sun May 22 09:24:04 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Sun, 22 May 2011 09:24:04 +0400
Subject: My first rewrite QUERY_STRING
In-Reply-To: <c50c4c6d2e5653978c341e8a3dc62427.NginxMailingListEnglish@forum.nginx.org>
References: <c50c4c6d2e5653978c341e8a3dc62427.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110522052404.GC65681@sysoev.ru>

On Sun, May 22, 2011 at 12:35:07AM -0400, yaz wrote:
> Hello 
> 
> here is a piece of htaccess i am stuck converting into nginx rewrite
> conventions : 
> 
> RewriteCond %{QUERY_STRING} ^sms_ss= [OR]
> RewriteCond %{QUERY_STRING} ^scvds=
> RewriteRule ^(.*)$ http://www.mysite.com/$1? [R=301,L]
> 
> 
> I am trying something like 
> 
>  if ($args ~ "sms_ss= (.*)") {
>         rewrite ^ /(.*)=$args? break;
>     }
> 
> Not working 

   if ($arg_ss) {
       rewrite  ^  $uri?  permanent;;
   }

   if ($arg_scvds) {
       rewrite  ^  $uri?  permanent;;
   }


-- 
Igor Sysoev


From nginx-forum at nginx.us  Sun May 22 09:57:41 2011
From: nginx-forum at nginx.us (yaz)
Date: Sun, 22 May 2011 01:57:41 -0400
Subject: My first rewrite QUERY_STRING
In-Reply-To: <c50c4c6d2e5653978c341e8a3dc62427.NginxMailingListEnglish@forum.nginx.org>
References: <c50c4c6d2e5653978c341e8a3dc62427.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <227bf7a88a8554edb8150998b1b6c09d.NginxMailingListEnglish@forum.nginx.org>

thanks igor, this is working. I will read the doc for better
understanding of rewriting with nginx

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199820,199832#msg-199832



From adam at lionite.com  Sun May 22 11:50:51 2011
From: adam at lionite.com (Adam Benayoun)
Date: Sun, 22 May 2011 10:50:51 +0300
Subject: simple rewrite rules
In-Reply-To: <20110522052155.GB65681@sysoev.ru>
References: <BANLkTi=gjCwvH6UkYdZawm2JvZbNXiWQdg@mail.gmail.com>
	<20110522052155.GB65681@sysoev.ru>
Message-ID: <BANLkTikHUbhOpAP=egEiPu+=w3NNbKWuAw@mail.gmail.com>

Igor,
Thanks for the pointer however there is one last thing I'm not sure how to
do.
Basically my application is built on top of a framework meaning that
everything should go through index.php and then the routers configured
appropriately should serve you different templates and controllers based on
the REQUEST_URI.

Basically when I try to access http://www.domain.com/category - it should
pass to index.php the request_uri /browse/category and then serve the
appropriate output.
This is how my config look like, let me know if that make sense and what
should I add to mimic that request_uri through index.php

Cheers,

server {
        listen *:80;
        server_name www.domain.com;
        access_log /var/www/domain/logs/access_log main;
        error_log /var/www/domain/logs/error_log debug;
        root   /var/www/domain/html/http;
        index index.php;

        if ($http_host !~ "^www\.domain\.com$") {
                rewrite  ^(.*)    http://www.domain.com$1 permanent;
        }

        location / {
                try_files $uri $uri/ /index.php?$args;
        }
        location /browse {
                #rewrite ^/browse/(.*) http://www.domain.com/$1 permanent;
        }
        location /min {
                try_files $uri $uri/ /min/index.php?q=;
        }
        location /blog {
                try_files $uri $uri/ /blog/index.php?q=$1;
        }
        location /apc {
                try_files $uri $uri/ /apc.php$args;
        }

        location ~ \.php {
                include /etc/nginx/fastcgi_params;
                fastcgi_param SCRIPT_FILENAME
$document_root$fastcgi_script_name;
                fastcgi_param PATH_INFO $fastcgi_script_name;
                fastcgi_param SERVER_NAME $http_host;
                fastcgi_pass 127.0.0.1:9000;
        }

        location ~* ^.+\.(ht|svn)$ {
                deny  all;
        }

        # Static files location
        location ~*
^.+\.(jpg|jpeg|gif|png|ico|css|zip|tgz|gz|rar|bz2|doc|xls|exe|pdf|ppt|txt|tar|mid|midi|wav|bmp|rtf|js)$
{
            expires max;
        }
}




On Sun, May 22, 2011 at 08:21, Igor Sysoev <igor at sysoev.ru> wrote:

> On Sun, May 22, 2011 at 03:41:59AM +0300, Adam Benayoun wrote:
> > Hello everyone,
> > I'm trying to setup some basic rewrite rules in nginx (I'm sorry if I may
> be
> > asking obvious question but I moved my servers from apache to nginx and I
> > can't figure out how to write simple rewrite rules in nginx).
> >
> > This is what I am trying to achieve:
> >
> > I have a URL with the following arguments:
> >
> > http://www.domain.com/browse/category
> >
> > What I want to achieve is:
> > - When someone try to reach http://www.domain.com/category - it will
> serve
> > the url http://www.domain.com/browse/category internally (no 301
> redirect).
> > - When someone will try to reach http://www.domain.com/browse/categoryit
> > will redirect externally (301) to http://www.domain.com/category
> >
> > I've added 2 rewriterule at first but somehow only the first rule works,
> the
> > second won't work:
> >
> >         rewrite ^/browse/(.*) http://www.domain.com/$1 permanent;
> >         rewrite ^/(category1|category2|category3|category4)$
>  /browse/$1
> > last;
> >
> > Thanks (and sorry for asking a somewhat newbie question!)
> >
> > p.s: yes, I've read this http://wiki.nginx.org/HttpRewriteModule
>
> You should not read anyting about rewrite module. It's better to forget
> about it. You should think using locations.
>
>     location /category1/
>         #depending on processing method:
>
>         # FastCGI
>         fastcgi_pass    backend;
>         fatscgi_param   SCRIPT_FILENMAE   /path/to/browse;
>
>         # proxying
>         proxy_pass    http://backend/browse/
>
>         # static files
>         root   /path/to/browse;
>     }
>
>     location /browse/ {
>         rewrite ^/browse/(.*)  /$1  permanent;
>     }
>
>
> --
> Igor Sysoev
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110522/74587710/attachment-0001.html>

From nginx-forum at nginx.us  Sun May 22 18:56:51 2011
From: nginx-forum at nginx.us (benseb)
Date: Sun, 22 May 2011 10:56:51 -0400
Subject: Sanity check of my config - is it secure?
Message-ID: <3ce420be0701a81ea969389d395b6a49.NginxMailingListEnglish@forum.nginx.org>

Hi

We had Nginx setup on a domain serving static content
(images[0-9].domain.com). Recently we've made our main domain also use
the same Nginx installation and proxy any php requests to Apache.

All is working fine. We've set it to serve all static content from our
images.domain.com domains, and it will also catch any static content
served from the main www.domain.com as well, before finally passing any
dynamic php requests through to Apache.

What I want to check is that there are no problems with the
configuration that would let someone view the source of our PHP pages or
access restricted 'admin' directories, etc. I've not configured NGinx
before so would appreciate some help checking my config?


user  www;
worker_processes  6;

error_log  /var/log/nginx-error.log;
#error_log  logs/error.log  notice;
#error_log  logs/error.log  info;

#pid        logs/nginx.pid;


events {
    worker_connections  1024;
}


http {
    include       mime.types;
    default_type  application/octet-stream;
    #access_log  logs/access.log  main;

    sendfile        on;
    #tcp_nopush     on;

     ## Timeouts
        client_body_timeout   10;
        client_header_timeout 10;
        keepalive_timeout     5 5;
        send_timeout          10;

    gzip  on;
#  gzip_static on;
    gzip_comp_level   5;
    gzip_http_version 1.0;
    gzip_min_length   1000;
    gzip_types        text/plain text/css image/x-icon
application/x-javascript;

    server{
                listen 80 default_server; #all other hosts deny
                server_name _;
                return 444;
    }

 server {  #IMAGES DOMAIN
        listen       x.x.x.x:80;
        server_name  images.x.com images1.x.com images2.x.com stat

        server_tokens off;
        gzip_comp_level 9;
        if ($request_method !~ ^(GET|HEAD)$ ) {
         return 444;
        }



        # Only serve these locations
        location ^~ /images/folder1/{

         if ($request_uri ~*
(^\/|\.ico|\.css|\.js|\.swf|\.gif|\.jp?g|\.png)$ ) {
          break;
        }

                root /home/vhosts/x.com/httpdocs;
                expires 7d;

        }
        location ^~ /images/folder2/{
               root /home/vhosts/x.com/httpdocs;
                expires 7d;

        }

         location ^~/images/ {
               root /home/vhosts/x.com/httpdocs;


        }


         location ^~ /applets/{
               root /home/vhosts/x.com/httpdocs;
               expires max;
        }
        location ^~ /css/{
               root /home/vhosts/x.com/httpdocs;
               expires max;
        }


        location ~* \.(ico|css|js|swf|gif|jp?g|png)$ {
                root /home/vhosts/x.com/httpdocs;
              
        }
        location ~* \.(php|html){
                return 444; #should not happen, but just incase
        }
                if ($host !~
^(images.x.com|images1.x.com|images2.x.com)){
                        return 444;
        }
        location ^~/admin/ {
                deny all;


        }



    }

server {   #MAIN DOMAIN - REDIRECT TO www.
        listen x.x.x.x:80;
        server_name x.com;
        rewrite ^(.*) http://www.x.com$1 permanent;

        }
 server { #MAIN DOMAIN
        listen       x.x.x.x:80;
        server_name www.x.com;

        #access_log /var/log/nginx.access.log;
        error_log /var/log/nginx.error.log;

         location /nginx_status {

            stub_status on;
            access_log   off;
           allow x.x.x.x/24;
            deny all;
        }

        location ^~ /images/folder1/{
               root /home/vhosts/x.com/httpdocs;
                expires 7d;

        }
        location ^~ /images/folder2/{
               root /home/vhosts/x.com/httpdocs;
                expires 7d;

        }
    

         location ^~/images/ {
               root /home/vhosts/x.com/httpdocs;

        }


         location ^~ /applets/{
               root /home/vhosts/x.com/httpdocs;
               expires max;
        }
        location ^~ /css/{
               root /home/vhosts/x.com/httpdocs;
               expires max;
        }


        location ~* \.(ico|css|js|swf|gif|jp?g|png)$ {
                root /home/vhosts/x.com/httpdocs;
                #expires max;
        }

        location / {
                proxy_pass http://127.0.0.1:8080;

                #CACHING
                #proxy_cache my-cache;
                #proxy_cache_valid 200 302 1m;
                #proxy_cache_valid 404 1m;

                 #proxy_redirect off;
                proxy_set_header Host $host;
               proxy_set_header X-Real-IP $remote_addr;
                proxy_set_header X-Forwarded-For
$proxy_add_x_forwarded_for;

                client_max_body_size 200m; #max upload size

                client_body_buffer_size 128k;
                proxy_connect_timeout 900;
                proxy_send_timeout 900;
                proxy_read_timeout 900;
                #proxy_buffer_size 4k;
                #proxy_buffers 4 32k;
                #proxy_busy_buffers_size 64k;
                #proxy_temp_file_write_size 64k;

        }
}

}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199902,199902#msg-199902



From nunomagalhaes at eu.ipp.pt  Sun May 22 20:48:42 2011
From: nunomagalhaes at eu.ipp.pt (=?UTF-8?Q?Nuno_Magalh=C3=A3es?=)
Date: Sun, 22 May 2011 17:48:42 +0100
Subject: Sanity check of my config - is it secure?
In-Reply-To: <3ce420be0701a81ea969389d395b6a49.NginxMailingListEnglish@forum.nginx.org>
References: <3ce420be0701a81ea969389d395b6a49.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTimOiuBPh7fYqMD+Cd1XKXskwJTAQA@mail.gmail.com>

Have a look at these:

http://wiki.nginx.org/Configuration
http://wiki.nginx.org/Modules
http://wiki.nginx.org/3rdPartyModules
https://calomel.org/nginx.html

And what version of nginx (latest is 1.0.2)? What OS?

On Sun, May 22, 2011 at 15:56, benseb <nginx-forum at nginx.us> wrote:

> ? ?gzip ?on;
> # ?gzip_static on;

If you're serving so much static content you may as well enable
gzip_static, although it's pointless for compressed images like jpg.

> ? ?gzip_types ? ? ? ?text/plain text/css image/x-icon

No html? Javascript?

> ? ? ? ?server_tokens off;

Won't disable the Server: header. There's a module that can do that.

> ? ? ? ? if ($request_uri ~*
> (^\/|\.ico|\.css|\.js|\.swf|\.gif|\.jp?g|\.png)$ ) {

You can probably replace this if-statement with a regex.

> ? ? ? ? location ^~ /applets/{
> ? ? ? ? ? ? ? root /home/vhosts/x.com/httpdocs;
> ? ? ? ? ? ? ? expires max;
> ? ? ? ?}
> ? ? ? ?location ^~ /css/{
> ? ? ? ? ? ? ? root /home/vhosts/x.com/httpdocs;
> ? ? ? ? ? ? ? expires max;
> ? ? ? ?}

And merge these locations (as well as others) perhaps? Something like
^~ /(applets|css)/

Just a few hints, like i said i'm no expert.

-- 
Mars 2 Stay!
http://xkcd.com/801/
/etc


From nginx-forum at nginx.us  Sun May 22 21:24:23 2011
From: nginx-forum at nginx.us (benseb)
Date: Sun, 22 May 2011 13:24:23 -0400
Subject: Sanity check of my config - is it secure?
In-Reply-To: <BANLkTimOiuBPh7fYqMD+Cd1XKXskwJTAQA@mail.gmail.com>
References: <BANLkTimOiuBPh7fYqMD+Cd1XKXskwJTAQA@mail.gmail.com>
Message-ID: <5542112b2862b2882ae3cbf516e14c16.NginxMailingListEnglish@forum.nginx.org>

Thanks for those tips. I'll have a read around those bits

My main concern is that by combining different locations, etc i may have
accidentally exposed a security issue - I need to ensure everything
that's not an image/css file/javascript file etc is either denied or
proxied to apache.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199902,199931#msg-199931



From justin at specialbusservice.com  Sun May 22 21:36:47 2011
From: justin at specialbusservice.com (Justin Cormack)
Date: Sun, 22 May 2011 18:36:47 +0100
Subject: Sanity check of my config - is it secure?
In-Reply-To: <5542112b2862b2882ae3cbf516e14c16.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTimOiuBPh7fYqMD+Cd1XKXskwJTAQA@mail.gmail.com>
	<5542112b2862b2882ae3cbf516e14c16.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <1306085807.3761.36.camel@mackerel>

On Sun, 2011-05-22 at 13:24 -0400, benseb wrote:
> Thanks for those tips. I'll have a read around those bits
> 
> My main concern is that by combining different locations, etc i may have
> accidentally exposed a security issue - I need to ensure everything
> that's not an image/css file/javascript file etc is either denied or
> proxied to apache.

You should write some tests! It is much better to use some well thought
out tests in addition to studying the config.

Justin




From nginx-forum at nginx.us  Sun May 22 22:37:46 2011
From: nginx-forum at nginx.us (benseb)
Date: Sun, 22 May 2011 14:37:46 -0400
Subject: Sanity check of my config - is it secure?
In-Reply-To: <1306085807.3761.36.camel@mackerel>
References: <1306085807.3761.36.camel@mackerel>
Message-ID: <a84874a004ee6b27f14c0b8050e43f93.NginxMailingListEnglish@forum.nginx.org>

For our static subdomain (which should only serve static content) I'd
like to add a whitelist of content - so if the filename doesn't match
jpg/png/gif/css/js etc it will return a 444

What's the correct syntax to do this:


location !~* \.(?:ico|css|js|gif|jpe?g|png)$ {
               #dont serve any other files
               return 444;
       }

The above doesnt work as !~* isn't allowed. What the best way to say
"not one of the following extensions". All the examples in the
documenation are the opposite?

Cheers

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199902,199950#msg-199950



From nginx-forum at nginx.us  Sun May 22 23:38:43 2011
From: nginx-forum at nginx.us (M.A.G)
Date: Sun, 22 May 2011 15:38:43 -0400
Subject: Nginx Vs Nginx Performance [Versions Battle]
Message-ID: <bf580a4e1661fe029db08aeac5294167.NginxMailingListEnglish@forum.nginx.org>

Hi All,
I would like to know if it's makes any difference in term of using Nginx
versions. Did anyone had any experience with that? Did anyone do any
benchmarks about that?
Looking for answers

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199956,199956#msg-199956



From sgsfak at gmail.com  Mon May 23 01:20:55 2011
From: sgsfak at gmail.com (Stelianos G. Sfakianakis)
Date: Mon, 23 May 2011 00:20:55 +0300
Subject: another rewrite/try files issue
Message-ID: <BANLkTin+0O-3oGFyZC=BusgP4Bp_7ozVMg@mail.gmail.com>

Hello,

I am having the following problem:

 - I want to have a "virtual" url e.g. "/myapp" to have a totally
different document root e.g. "/home/stelios/myapp" (instead of the
default '/var/www')
 - locations of the type "/myapp/css/print.css" should respond with
the specific file at the  "/home/stelios/myapp" root (i.e. with
"/home/stelios/myapp/css/print.css"
 - if the the file requested does not exist, then the "index.php"
should be called via FastCGI with args "?u=$uri" (e.g.
"/myapp/users/12" should be rewritten to
"/home/stelios/myapp/index.php?u=users/12")

I have tried the following:

location ~ /myapp/(.*) {
                root /home/stelios/myapp;
                try_files /$1 /$1/ /index.php?u=$1;
}
# and the "default" PHP/FastCGI
location ~ \.php$ {
                fastcgi_pass 127.0.0.1:9000;
                fastcgi_index index.php;
                fastcgi_param SCRIPT_FILENAME
$document_root$fastcgi_script_name;
                include fastcgi_params;
}


but it seems that it doesn't work for the non files. In particular it
seems that after the internal redirect to index.php, the root is
overwritten and instead of calling my  /home/stelios/myapp/index.php
the one at the default root ("/var/www/index.php") is beeing called.

Do you know what's wrong? And possibl another way to achieve what I want?

Thank you very much in advance!

P.S. I am using nginx-0.8.4 on Ubuntu

St.


From nginx-forum at nginx.us  Mon May 23 02:06:01 2011
From: nginx-forum at nginx.us (brianmercer)
Date: Sun, 22 May 2011 18:06:01 -0400
Subject: another rewrite/try files issue
In-Reply-To: <BANLkTin+0O-3oGFyZC=BusgP4Bp_7ozVMg@mail.gmail.com>
References: <BANLkTin+0O-3oGFyZC=BusgP4Bp_7ozVMg@mail.gmail.com>
Message-ID: <ca182fbfac7cd8c4bf2a0246c2a2ce8c.NginxMailingListEnglish@forum.nginx.org>

Stelianos G. Sfakianakis Wrote:
-------------------------------------------------------
> Hello,
> 
> I am having the following problem:
> 
>  - I want to have a "virtual" url e.g. "/myapp" to
> have a totally
> different document root e.g. "/home/stelios/myapp"
> (instead of the
> default '/var/www')
>  - locations of the type "/myapp/css/print.css"
> should respond with
> the specific file at the  "/home/stelios/myapp"
> root (i.e. with
> "/home/stelios/myapp/css/print.css"
>  - if the the file requested does not exist, then
> the "index.php"
> should be called via FastCGI with args "?u=$uri"
> (e.g.
> "/myapp/users/12" should be rewritten to
> "/home/stelios/myapp/index.php?u=users/12")
> 
> I have tried the following:
> 
> location ~ /myapp/(.*) {
>                 root /home/stelios/myapp;
>                 try_files /$1 /$1/
> /index.php?u=$1;
> }
> # and the "default" PHP/FastCGI
> location ~ \.php$ {
>                 fastcgi_pass 127.0.0.1:9000;
>                 fastcgi_index index.php;
>                 fastcgi_param SCRIPT_FILENAME
> $document_root$fastcgi_script_name;
>                 include fastcgi_params;
> }
> 
> 
> but it seems that it doesn't work for the non
> files. In particular it
> seems that after the internal redirect to
> index.php, the root is
> overwritten and instead of calling my 
> /home/stelios/myapp/index.php
> the one at the default root ("/var/www/index.php")
> is beeing called.
> 
> Do you know what's wrong? And possibl another way
> to achieve what I want?
> 
> Thank you very much in advance!
> 
> P.S. I am using nginx-0.8.4 on Ubuntu
> 
> St.
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx

Possibly this:

location /myapp/ {
  root /home/stelios;
  try_files $uri @fallback;
}

# and the "default" PHP/FastCGI
location @fallback {
  fastcgi_pass 127.0.0.1:9000;
  fastcgi_param SCRIPT_FILENAME /home/stelios/myapp/index.php;
  include fastcgi_params;
}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199965,199970#msg-199970



From nginx-forum at nginx.us  Mon May 23 04:31:59 2011
From: nginx-forum at nginx.us (ltss)
Date: Sun, 22 May 2011 20:31:59 -0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <6856e4d76e54e6ffbc66eaa7c05442ce.NginxMailingListEnglish@forum.nginx.org>
References: <c0c6df7135604bcee44a098e5b896a30@ruby-forum.com>
	<e00b353ec548b2c91b169c5b8ccaaa9c.NginxMailingListEnglish@forum.nginx.org>
	<6856e4d76e54e6ffbc66eaa7c05442ce.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4afe9fa1984e2ca74c7636d2a379d417.NginxMailingListEnglish@forum.nginx.org>

Thanks that work really fine ;)

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,23868,199984#msg-199984



From sgsfak at gmail.com  Mon May 23 10:53:08 2011
From: sgsfak at gmail.com (Stelios Sfakianakis)
Date: Mon, 23 May 2011 09:53:08 +0300
Subject: another rewrite/try files issue
In-Reply-To: <ca182fbfac7cd8c4bf2a0246c2a2ce8c.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTin+0O-3oGFyZC=BusgP4Bp_7ozVMg@mail.gmail.com>
	<ca182fbfac7cd8c4bf2a0246c2a2ce8c.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4DDA0454.5080101@gmail.com>

On 23/5/2011 1:06 ??, brianmercer wrote:
> Possibly this:
>
> location /myapp/ {
>    root /home/stelios;
>    try_files $uri @fallback;
> }
>
> # and the "default" PHP/FastCGI
> location @fallback {
>    fastcgi_pass 127.0.0.1:9000;
>    fastcgi_param SCRIPT_FILENAME /home/stelios/myapp/index.php;
>    include fastcgi_params;
> }
>

Thanks Brian, that *almost* works (although I don't like repeating [most 
of] the fastcgi config..)

But I also need to pass *part* of the original path as a query string. I 
tried to following :

location @fallback {
       fastcgi_pass 127.0.0.1:9000;
       include fastcgi_params;
       fastcgi_param SCRIPT_FILENAME /home/stelios/myapp/index.php;
       fastcgi_param  QUERY_STRING  u=$uri&$args;
}


which works but now for the url ".../myapp/users" the script gets a 
query string like this: "u=/myapp/users" whereas I want it to be 
"u=/users" (i.e. remove the "base" path). Is there any simple solution 
to that? (e.g. can we pass parameters to a named location?)

Thanks!
Stelios



From agentzh at gmail.com  Mon May 23 12:46:32 2011
From: agentzh at gmail.com (agentzh)
Date: Mon, 23 May 2011 16:46:32 +0800
Subject: lua location capture
In-Reply-To: <EFC5A19DBF3A2E489904733B467D990B125375@office-server2.m247.local>
References: <EFC5A19DBF3A2E489904733B467D990B12493E@office-server2.m247.local>
	<BANLkTi=_eN=EzKF-J9=cFgu6qd29GVOwuA@mail.gmail.com>
	<EFC5A19DBF3A2E489904733B467D990B125375@office-server2.m247.local>
Message-ID: <BANLkTimZLH2ZL4NFA1PO8BBLno_07vNRpw@mail.gmail.com>

On Fri, May 20, 2011 at 7:32 PM, Richard Kearsley
<Richard.Kearsley at m247.com> wrote:
> I traced the issue with the header's not being re-sent
>
> ngx.location.capture("/some-location", { method = ngx.HTTP_HEAD })
> they are not re-sent
>
> but
> ngx.location.capture("/some-location", { method = ngx.HTTP_GET })
> they are sent
>
> I don't know if this is correct behavior (it may well be)
> I got around it by changing the headers I needed into args then re-creating them at /some-location :(
>

This issue has been proven to be caused by two separate bugs. I've
already fixed them in ngx_lua's master branch:

    http://bit.ly/mU1Z29

Could you please try it out?

Best,
-agentzh


From roberto at unbit.it  Mon May 23 18:18:30 2011
From: roberto at unbit.it (Roberto De Ioris)
Date: Mon, 23 May 2011 16:18:30 +0200 (CEST)
Subject: [PATH] support for UNIX socket in abstract namespace
Message-ID: <9bdc098f7db74f8cfdc102938169f26a.squirrel@manage.unbit.it>

(Warning this is Linux only)

Hi all, the attached path adds support for unix socket in abstract namespace.
They are special sockets without filesystem correspondence (so you can use
them without thinking about permissions or in chroot). In netstat they are
reported with a '@' prefix.

For example (using uWSGI):

uwsgi -s @funnysock

netstat -l

unix  2      [ ACC ]     STREAM     LISTENING     23813    @funnysock

After applying the patch you can connect to it with

uwsgi_pass unix:@funnysock;

I hope it can be useful

-- 
Roberto De Ioris
http://unbit.it
-------------- next part --------------
A non-text attachment was scrubbed...
Name: unix_abstract.patch
Type: application/octet-stream
Size: 1049 bytes
Desc: not available
URL: <http://nginx.org/pipermail/nginx/attachments/20110523/f704fe53/attachment.obj>

From david at styleflare.com  Mon May 23 20:13:27 2011
From: david at styleflare.com (David J.)
Date: Mon, 23 May 2011 12:13:27 -0400
Subject: Basic Auth - MySQL/PgSQL Backend.
Message-ID: <4DDA87A7.8020200@styleflare.com>

I was wondering what options are available for using BasicAuth using 
Either MySQL or PostGres backend to store users.





From igor at sysoev.ru  Mon May 23 20:35:41 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Mon, 23 May 2011 20:35:41 +0400
Subject: [PATH] support for UNIX socket in abstract namespace
In-Reply-To: <9bdc098f7db74f8cfdc102938169f26a.squirrel@manage.unbit.it>
References: <9bdc098f7db74f8cfdc102938169f26a.squirrel@manage.unbit.it>
Message-ID: <20110523163541.GA39527@sysoev.ru>

On Mon, May 23, 2011 at 04:18:30PM +0200, Roberto De Ioris wrote:
> (Warning this is Linux only)
> 
> Hi all, the attached path adds support for unix socket in abstract namespace.
> They are special sockets without filesystem correspondence (so you can use
> them without thinking about permissions or in chroot). In netstat they are
> reported with a '@' prefix.
> 
> For example (using uWSGI):
> 
> uwsgi -s @funnysock
> 
> netstat -l
> 
> unix  2      [ ACC ]     STREAM     LISTENING     23813    @funnysock
> 
> After applying the patch you can connect to it with
> 
> uwsgi_pass unix:@funnysock;
> 
> I hope it can be useful

Thank you for the feature. I think it's better to allow nginx configuration
parser to support "\0" as binary 0. Then this socket may be set as

uwsgi_pass   unix:\0funnysock;


-- 
Igor Sysoev


From piotr.sikora at frickle.com  Mon May 23 20:49:37 2011
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Mon, 23 May 2011 18:49:37 +0200
Subject: Basic Auth - MySQL/PgSQL Backend.
In-Reply-To: <4DDA87A7.8020200@styleflare.com>
References: <4DDA87A7.8020200@styleflare.com>
Message-ID: <07C71B337A6E42D8AB32D3578CD2CE03@Desktop>

Hi,

> I was wondering what options are available for using BasicAuth using 
> Either MySQL or PostGres backend to store users.

You can do this with ngx_http_auth_request [1] with either ngx_drizzle [2] 
(for Drizzle/MariaDB/MySQL/Percona) or ngx_postgres [3] (for PostgreSQL).

Working example for ngx_postgres is available in its README (Sample 
configuration #4).

[1] http://mdounin.ru/hg/ngx_http_auth_request_module/
[2] https://github.com/chaoslawful/drizzle-nginx-module
[3] https://github.com/FRiCKLE/ngx_postgres

Best regards,
Piotr Sikora < piotr.sikora at frickle.com >



From david at styleflare.com  Mon May 23 20:55:38 2011
From: david at styleflare.com (David J.)
Date: Mon, 23 May 2011 12:55:38 -0400
Subject: Basic Auth - MySQL/PgSQL Backend.
In-Reply-To: <07C71B337A6E42D8AB32D3578CD2CE03@Desktop>
References: <4DDA87A7.8020200@styleflare.com>
	<07C71B337A6E42D8AB32D3578CD2CE03@Desktop>
Message-ID: <4DDA918A.3010701@styleflare.com>

Thanks.

I didnt see the example in the README?
  Sample #4?

Thanks Again.

David.

On 5/23/11 12:49 PM, Piotr Sikora wrote:
> Hi,
>
>> I was wondering what options are available for using BasicAuth using 
>> Either MySQL or PostGres backend to store users.
>
> You can do this with ngx_http_auth_request [1] with either ngx_drizzle 
> [2] (for Drizzle/MariaDB/MySQL/Percona) or ngx_postgres [3] (for 
> PostgreSQL).
>
> Working example for ngx_postgres is available in its README (Sample 
> configuration #4).
>
> [1] http://mdounin.ru/hg/ngx_http_auth_request_module/
> [2] https://github.com/chaoslawful/drizzle-nginx-module
> [3] https://github.com/FRiCKLE/ngx_postgres
>
> Best regards,
> Piotr Sikora < piotr.sikora at frickle.com >
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>



From piotr.sikora at frickle.com  Mon May 23 21:06:04 2011
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Mon, 23 May 2011 19:06:04 +0200
Subject: Basic Auth - MySQL/PgSQL Backend.
In-Reply-To: <4DDA918A.3010701@styleflare.com>
References: <4DDA87A7.8020200@styleflare.com><07C71B337A6E42D8AB32D3578CD2CE03@Desktop>
	<4DDA918A.3010701@styleflare.com>
Message-ID: <31FCBEFC2F9641ED9451C2CBE0120518@Desktop>

Hi,

> I didnt see the example in the README?
>  Sample #4?

Then look again, it's there.

Best regards,
Piotr Sikora < piotr.sikora at frickle.com >



From david at styleflare.com  Mon May 23 21:09:36 2011
From: david at styleflare.com (David J.)
Date: Mon, 23 May 2011 13:09:36 -0400
Subject: Basic Auth - MySQL/PgSQL Backend.
In-Reply-To: <31FCBEFC2F9641ED9451C2CBE0120518@Desktop>
References: <4DDA87A7.8020200@styleflare.com><07C71B337A6E42D8AB32D3578CD2CE03@Desktop>	<4DDA918A.3010701@styleflare.com>
	<31FCBEFC2F9641ED9451C2CBE0120518@Desktop>
Message-ID: <4DDA94D0.5080908@styleflare.com>

I am not sure which readme you are referring to;

I see your name is  listed here;

https://github.com/chaoslawful/drizzle-nginx-module

Is this the file you are talking about?

Thanks.



On 5/23/11 1:06 PM, Piotr Sikora wrote:
> Hi,
>
>> I didnt see the example in the README?
>>  Sample #4?
>
> Then look again, it's there.
>
> Best regards,
> Piotr Sikora < piotr.sikora at frickle.com >
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>



From nginx-forum at nginx.us  Mon May 23 21:18:30 2011
From: nginx-forum at nginx.us (the_traveller)
Date: Mon, 23 May 2011 13:18:30 -0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <4afe9fa1984e2ca74c7636d2a379d417.NginxMailingListEnglish@forum.nginx.org>
References: <c0c6df7135604bcee44a098e5b896a30@ruby-forum.com>
	<e00b353ec548b2c91b169c5b8ccaaa9c.NginxMailingListEnglish@forum.nginx.org>
	<6856e4d76e54e6ffbc66eaa7c05442ce.NginxMailingListEnglish@forum.nginx.org>
	<4afe9fa1984e2ca74c7636d2a379d417.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <eaddf429fe128da4d46fcde97e1ddf24.NginxMailingListEnglish@forum.nginx.org>

Hi mr Igor 
I'm sory coz i toke alittle bit too long

so if i have to make any ngnix rule i have to make it like that by ex:

location /Joke/ {
fastcgi_pass ...
fastcgi_param SCRIPT_FILENAME /jokes.php;
fastcgi_param QUERY_STRING $args;
...
}

and the Original Apache rule is
RewriteRule ^Joke/.*     jokes.php       [QSA]

sposing that the script is in the public dir  of the ftp 
is my try right??

and also is the three dots in the second line
fastcgi_pass ...
and before the closing tag
...
}
 is this is right or i have to remove them ??

thanks alot for ur help
and if i have removed them

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,23868,200161#msg-200161



From piotr.sikora at frickle.com  Mon May 23 21:19:25 2011
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Mon, 23 May 2011 19:19:25 +0200
Subject: Basic Auth - MySQL/PgSQL Backend.
In-Reply-To: <4DDA94D0.5080908@styleflare.com>
References: <4DDA87A7.8020200@styleflare.com><07C71B337A6E42D8AB32D3578CD2CE03@Desktop>	<4DDA918A.3010701@styleflare.com><31FCBEFC2F9641ED9451C2CBE0120518@Desktop>
	<4DDA94D0.5080908@styleflare.com>
Message-ID: <3FB95AA462DD4243BE9671D18CA04BCE@Desktop>

Hi,

> I am not sure which readme you are referring to;

"Working example for ngx_postgres is available in its README (Sample 
configuration #4)." should be self-explanatory ;)

> I see your name is  listed here;
>
> https://github.com/chaoslawful/drizzle-nginx-module

I'm listed in the other one as well.

> Is this the file you are talking about?

Nope, I'm talking about ngx_postgres's one.

Best regards,
Piotr Sikora < piotr.sikora at frickle.com >



From david at styleflare.com  Mon May 23 21:22:16 2011
From: david at styleflare.com (David J.)
Date: Mon, 23 May 2011 13:22:16 -0400
Subject: Basic Auth - MySQL/PgSQL Backend.
In-Reply-To: <3FB95AA462DD4243BE9671D18CA04BCE@Desktop>
References: <4DDA87A7.8020200@styleflare.com><07C71B337A6E42D8AB32D3578CD2CE03@Desktop>	<4DDA918A.3010701@styleflare.com><31FCBEFC2F9641ED9451C2CBE0120518@Desktop>	<4DDA94D0.5080908@styleflare.com>
	<3FB95AA462DD4243BE9671D18CA04BCE@Desktop>
Message-ID: <4DDA97C8.3030601@styleflare.com>

Thanks.

I really appreciate it;

That (surprisingly!) was the only one I didnt look at.

Thanks again.


On 5/23/11 1:19 PM, Piotr Sikora wrote:
> Hi,
>
>> I am not sure which readme you are referring to;
>
> "Working example for ngx_postgres is available in its README (Sample 
> configuration #4)." should be self-explanatory ;)
>
>> I see your name is  listed here;
>>
>> https://github.com/chaoslawful/drizzle-nginx-module
>
> I'm listed in the other one as well.
>
>> Is this the file you are talking about?
>
> Nope, I'm talking about ngx_postgres's one.
>
> Best regards,
> Piotr Sikora < piotr.sikora at frickle.com >
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>



From roberto at unbit.it  Mon May 23 21:42:02 2011
From: roberto at unbit.it (Roberto De Ioris)
Date: Mon, 23 May 2011 19:42:02 +0200 (CEST)
Subject: [PATH] support for UNIX socket in abstract namespace
In-Reply-To: <20110523163541.GA39527@sysoev.ru>
References: <9bdc098f7db74f8cfdc102938169f26a.squirrel@manage.unbit.it>
	<20110523163541.GA39527@sysoev.ru>
Message-ID: <45113c3f74b15ed88b9bdeff9a2e7a20.squirrel@manage.unbit.it>


> On Mon, May 23, 2011 at 04:18:30PM +0200, Roberto De Ioris wrote:
>> (Warning this is Linux only)
>>
>> Hi all, the attached path adds support for unix socket in abstract
>> namespace.
>> They are special sockets without filesystem correspondence (so you can
>> use
>> them without thinking about permissions or in chroot). In netstat they
>> are
>> reported with a '@' prefix.
>>
>> For example (using uWSGI):
>>
>> uwsgi -s @funnysock
>>
>> netstat -l
>>
>> unix  2      [ ACC ]     STREAM     LISTENING     23813    @funnysock
>>
>> After applying the patch you can connect to it with
>>
>> uwsgi_pass unix:@funnysock;
>>
>> I hope it can be useful
>
> Thank you for the feature. I think it's better to allow nginx
> configuration
> parser to support "\0" as binary 0. Then this socket may be set as
>
> uwsgi_pass   unix:\0funnysock;
>

I fear it will require changes changes as 0 being the token delimeter in
config parsing :(


-- 
Roberto De Ioris
http://unbit.it


From roberto at unbit.it  Mon May 23 21:44:06 2011
From: roberto at unbit.it (Roberto De Ioris)
Date: Mon, 23 May 2011 19:44:06 +0200 (CEST)
Subject: [PATH] support for UNIX socket in abstract namespace
In-Reply-To: <45113c3f74b15ed88b9bdeff9a2e7a20.squirrel@manage.unbit.it>
References: <9bdc098f7db74f8cfdc102938169f26a.squirrel@manage.unbit.it>
	<20110523163541.GA39527@sysoev.ru>
	<45113c3f74b15ed88b9bdeff9a2e7a20.squirrel@manage.unbit.it>
Message-ID: <034aea7b75fecac7c87e6d9e229f6ad0.squirrel@manage.unbit.it>


>>> I hope it can be useful
>>
>> Thank you for the feature. I think it's better to allow nginx
>> configuration
>> parser to support "\0" as binary 0. Then this socket may be set as
>>
>> uwsgi_pass   unix:\0funnysock;
>>
>
> I fear it will require changes changes



i mean "heavy changes"



-- 
Roberto De Ioris
http://unbit.it


From igor at sysoev.ru  Mon May 23 21:48:56 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Mon, 23 May 2011 21:48:56 +0400
Subject: [PATH] support for UNIX socket in abstract namespace
In-Reply-To: <45113c3f74b15ed88b9bdeff9a2e7a20.squirrel@manage.unbit.it>
References: <9bdc098f7db74f8cfdc102938169f26a.squirrel@manage.unbit.it>
	<20110523163541.GA39527@sysoev.ru>
	<45113c3f74b15ed88b9bdeff9a2e7a20.squirrel@manage.unbit.it>
Message-ID: <20110523174856.GC39527@sysoev.ru>

On Mon, May 23, 2011 at 07:42:02PM +0200, Roberto De Ioris wrote:
> 
> > On Mon, May 23, 2011 at 04:18:30PM +0200, Roberto De Ioris wrote:
> >> (Warning this is Linux only)
> >>
> >> Hi all, the attached path adds support for unix socket in abstract
> >> namespace.
> >> They are special sockets without filesystem correspondence (so you can
> >> use
> >> them without thinking about permissions or in chroot). In netstat they
> >> are
> >> reported with a '@' prefix.
> >>
> >> For example (using uWSGI):
> >>
> >> uwsgi -s @funnysock
> >>
> >> netstat -l
> >>
> >> unix  2      [ ACC ]     STREAM     LISTENING     23813    @funnysock
> >>
> >> After applying the patch you can connect to it with
> >>
> >> uwsgi_pass unix:@funnysock;
> >>
> >> I hope it can be useful
> >
> > Thank you for the feature. I think it's better to allow nginx
> > configuration
> > parser to support "\0" as binary 0. Then this socket may be set as
> >
> > uwsgi_pass   unix:\0funnysock;
> >
> 
> I fear it will require changes changes as 0 being the token delimeter in
> config parsing :(

No, nginx does not use zero-trailing strings.


-- 
Igor Sysoev


From igor at sysoev.ru  Mon May 23 21:51:51 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Mon, 23 May 2011 21:51:51 +0400
Subject: [PATH] support for UNIX socket in abstract namespace
In-Reply-To: <45113c3f74b15ed88b9bdeff9a2e7a20.squirrel@manage.unbit.it>
References: <9bdc098f7db74f8cfdc102938169f26a.squirrel@manage.unbit.it>
	<20110523163541.GA39527@sysoev.ru>
	<45113c3f74b15ed88b9bdeff9a2e7a20.squirrel@manage.unbit.it>
Message-ID: <20110523175151.GD39527@sysoev.ru>

On Mon, May 23, 2011 at 07:42:02PM +0200, Roberto De Ioris wrote:
> 
> > On Mon, May 23, 2011 at 04:18:30PM +0200, Roberto De Ioris wrote:
> >> (Warning this is Linux only)
> >>
> >> Hi all, the attached path adds support for unix socket in abstract
> >> namespace.
> >> They are special sockets without filesystem correspondence (so you can
> >> use
> >> them without thinking about permissions or in chroot). In netstat they
> >> are
> >> reported with a '@' prefix.
> >>
> >> For example (using uWSGI):
> >>
> >> uwsgi -s @funnysock
> >>
> >> netstat -l
> >>
> >> unix  2      [ ACC ]     STREAM     LISTENING     23813    @funnysock
> >>
> >> After applying the patch you can connect to it with
> >>
> >> uwsgi_pass unix:@funnysock;
> >>
> >> I hope it can be useful
> >
> > Thank you for the feature. I think it's better to allow nginx
> > configuration
> > parser to support "\0" as binary 0. Then this socket may be set as
> >
> > uwsgi_pass   unix:\0funnysock;
> >
> 
> I fear it will require changes changes as 0 being the token delimeter in
> config parsing :(

Could you test the attached patch ?


-- 
Igor Sysoev
-------------- next part --------------
Index: src/core/ngx_conf_file.c
===================================================================
--- src/core/ngx_conf_file.c	(revision 3923)
+++ src/core/ngx_conf_file.c	(working copy)
@@ -714,6 +714,11 @@
                             *dst++ = '\n';
                             src += 2;
                             continue;
+
+                        case '0':
+                            *dst++ = '\0';
+                            src += 2;
+                            continue;
                         }
 
                     }

From nginx-forum at nginx.us  Mon May 23 22:02:43 2011
From: nginx-forum at nginx.us (brianmercer)
Date: Mon, 23 May 2011 14:02:43 -0400
Subject: another rewrite/try files issue
In-Reply-To: <4DDA0454.5080101@gmail.com>
References: <4DDA0454.5080101@gmail.com>
Message-ID: <64a6135089b22eb9bfcfafb93ef9ceba.NginxMailingListEnglish@forum.nginx.org>

Stelios Sfakianakis Wrote:
-------------------------------------------------------
> On 23/5/2011 1:06 ??, brianmercer wrote:
> > Possibly this:
> >
> > location /myapp/ {
> >    root /home/stelios;
> >    try_files $uri @fallback;
> > }
> >
> > # and the "default" PHP/FastCGI
> > location @fallback {
> >    fastcgi_pass 127.0.0.1:9000;
> >    fastcgi_param SCRIPT_FILENAME
> /home/stelios/myapp/index.php;
> >    include fastcgi_params;
> > }
> >
> 
> Thanks Brian, that *almost* works (although I
> don't like repeating [most 
> of] the fastcgi config..)
> 
> But I also need to pass *part* of the original
> path as a query string. I 
> tried to following :
> 
> location @fallback {
>        fastcgi_pass 127.0.0.1:9000;
>        include fastcgi_params;
>        fastcgi_param SCRIPT_FILENAME
> /home/stelios/myapp/index.php;
>        fastcgi_param  QUERY_STRING  u=$uri&$args;
> }
> 
> 
> which works but now for the url ".../myapp/users"
> the script gets a 
> query string like this: "u=/myapp/users" whereas I
> want it to be 
> "u=/users" (i.e. remove the "base" path). Is there
> any simple solution 
> to that? (e.g. can we pass parameters to a named
> location?)
> 
> Thanks!
> Stelios
> 
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx

Oh, I see.  Maybe this:

location ~ /myapp/(?<myquery>.*) {
root /home/stelios;
try_files $uri /myapp/index.php?u=/$myquery;
}

location = /myapp/index.php {
fastcgi_pass 127.0.0.1:9000;
fastcgi_param SCRIPT_FILENAME /home/stelios/myapp/index.php;
include fastcgi_params;
}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199965,200179#msg-200179



From roberto at unbit.it  Mon May 23 22:17:46 2011
From: roberto at unbit.it (Roberto De Ioris)
Date: Mon, 23 May 2011 20:17:46 +0200 (CEST)
Subject: [PATH] support for UNIX socket in abstract namespace
In-Reply-To: <20110523175151.GD39527@sysoev.ru>
References: <9bdc098f7db74f8cfdc102938169f26a.squirrel@manage.unbit.it>
	<20110523163541.GA39527@sysoev.ru>
	<45113c3f74b15ed88b9bdeff9a2e7a20.squirrel@manage.unbit.it>
	<20110523175151.GD39527@sysoev.ru>
Message-ID: <dd26a4b35a34020513bd2483afe1ba5e.squirrel@manage.unbit.it>


> On Mon, May 23, 2011 at 07:42:02PM +0200, Roberto De Ioris wrote:
>>
>> > On Mon, May 23, 2011 at 04:18:30PM +0200, Roberto De Ioris wrote:
>> >> (Warning this is Linux only)
>> >>
>> >> Hi all, the attached path adds support for unix socket in abstract
>> >> namespace.
>> >> They are special sockets without filesystem correspondence (so you
>> can
>> >> use
>> >> them without thinking about permissions or in chroot). In netstat
>> they
>> >> are
>> >> reported with a '@' prefix.
>> >>
>> >> For example (using uWSGI):
>> >>
>> >> uwsgi -s @funnysock
>> >>
>> >> netstat -l
>> >>
>> >> unix  2      [ ACC ]     STREAM     LISTENING     23813    @funnysock
>> >>
>> >> After applying the patch you can connect to it with
>> >>
>> >> uwsgi_pass unix:@funnysock;
>> >>
>> >> I hope it can be useful
>> >
>> > Thank you for the feature. I think it's better to allow nginx
>> > configuration
>> > parser to support "\0" as binary 0. Then this socket may be set as
>> >
>> > uwsgi_pass   unix:\0funnysock;
>> >
>>
>> I fear it will require changes changes as 0 being the token delimeter in
>> config parsing :(
>
> Could you test the attached patch ?


This is the same code i made after you suggestion, but in the url parsing
code (ngx_parse_unix_domain_url), ngx_cpystrn() is used and it takes '\0'
in account.

Changing ngx_cpystrn to not use \0 fixes the problem. Probably
substituting it with ngx_memcpy is the way to go.

Another thing to do is "fixing"

u->addrs[0].socklen = sizeof(struct sockaddr_un);

to

u->addrs[0].socklen = sizeof(saun->sun_family)+len;



-- 
Roberto De Ioris
http://unbit.it


From igor at sysoev.ru  Mon May 23 22:37:26 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Mon, 23 May 2011 22:37:26 +0400
Subject: [PATH] support for UNIX socket in abstract namespace
In-Reply-To: <dd26a4b35a34020513bd2483afe1ba5e.squirrel@manage.unbit.it>
References: <9bdc098f7db74f8cfdc102938169f26a.squirrel@manage.unbit.it>
	<20110523163541.GA39527@sysoev.ru>
	<45113c3f74b15ed88b9bdeff9a2e7a20.squirrel@manage.unbit.it>
	<20110523175151.GD39527@sysoev.ru>
	<dd26a4b35a34020513bd2483afe1ba5e.squirrel@manage.unbit.it>
Message-ID: <20110523183726.GE39527@sysoev.ru>

On Mon, May 23, 2011 at 08:17:46PM +0200, Roberto De Ioris wrote:
> 
> > On Mon, May 23, 2011 at 07:42:02PM +0200, Roberto De Ioris wrote:
> >>
> >> > On Mon, May 23, 2011 at 04:18:30PM +0200, Roberto De Ioris wrote:
>
> This is the same code i made after you suggestion, but in the url parsing
> code (ngx_parse_unix_domain_url), ngx_cpystrn() is used and it takes '\0'
> in account.
> 
> Changing ngx_cpystrn to not use \0 fixes the problem. Probably
> substituting it with ngx_memcpy is the way to go.
> 
> Another thing to do is "fixing"
> 
> u->addrs[0].socklen = sizeof(struct sockaddr_un);
> 
> to
> 
> u->addrs[0].socklen = sizeof(saun->sun_family)+len;

nxg_memcpy should work here.


-- 
Igor Sysoev


From sgsfak at gmail.com  Mon May 23 23:31:31 2011
From: sgsfak at gmail.com (Stelianos G. Sfakianakis)
Date: Mon, 23 May 2011 22:31:31 +0300
Subject: another rewrite/try files issue
In-Reply-To: <64a6135089b22eb9bfcfafb93ef9ceba.NginxMailingListEnglish@forum.nginx.org>
References: <4DDA0454.5080101@gmail.com>
	<64a6135089b22eb9bfcfafb93ef9ceba.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTikxstF3+osH+bU-asLD28KzVP4BgA@mail.gmail.com>

2011/5/23 brianmercer <nginx-forum at nginx.us>:
> Oh, I see. ?Maybe this:
>
> location ~ /myapp/(?<myquery>.*) {
> root /home/stelios;
> try_files $uri /myapp/index.php?u=/$myquery;
> }
>
> location = /myapp/index.php {
> fastcgi_pass 127.0.0.1:9000;
> fastcgi_param SCRIPT_FILENAME /home/stelios/myapp/index.php;
> include fastcgi_params;
> }

Thanks again!

I finally did the following which seems to work just fine:

location ~ /myapp/(.*) {
     root /home/stelios/myapp;
     set $args u=/$1&$args;
     try_files /$1 @fallback;
}
location @fallback {
     root /home/stelios/myapp;
    fastcgi_pass 127.0.0.1:9000;
    include fastcgi_params;
    fastcgi_param SCRIPT_FILENAME $document_root/index.php;
}

The "set $args ..." directive augments the QUERY_STRING with the
original uri that is kept to the subsequent fastcgi call.

Best regards
Stelios


From igor at sysoev.ru  Tue May 24 00:02:27 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 24 May 2011 00:02:27 +0400
Subject: another rewrite/try files issue
In-Reply-To: <BANLkTin+0O-3oGFyZC=BusgP4Bp_7ozVMg@mail.gmail.com>
References: <BANLkTin+0O-3oGFyZC=BusgP4Bp_7ozVMg@mail.gmail.com>
Message-ID: <20110523200226.GG39527@sysoev.ru>

On Mon, May 23, 2011 at 12:20:55AM +0300, Stelianos G. Sfakianakis wrote:
> Hello,
> 
> I am having the following problem:
> 
>  - I want to have a "virtual" url e.g. "/myapp" to have a totally
> different document root e.g. "/home/stelios/myapp" (instead of the
> default '/var/www')
>  - locations of the type "/myapp/css/print.css" should respond with
> the specific file at the  "/home/stelios/myapp" root (i.e. with
> "/home/stelios/myapp/css/print.css"
>  - if the the file requested does not exist, then the "index.php"
> should be called via FastCGI with args "?u=$uri" (e.g.
> "/myapp/users/12" should be rewritten to
> "/home/stelios/myapp/index.php?u=users/12")
> 
> I have tried the following:
> 
> location ~ /myapp/(.*) {
>                 root /home/stelios/myapp;
>                 try_files /$1 /$1/ /index.php?u=$1;
> }
> # and the "default" PHP/FastCGI
> location ~ \.php$ {
>                 fastcgi_pass 127.0.0.1:9000;
>                 fastcgi_index index.php;
>                 fastcgi_param SCRIPT_FILENAME
> $document_root$fastcgi_script_name;
>                 include fastcgi_params;
> }
> 
> 
> but it seems that it doesn't work for the non files. In particular it
> seems that after the internal redirect to index.php, the root is
> overwritten and instead of calling my  /home/stelios/myapp/index.php
> the one at the default root ("/var/www/index.php") is beeing called.
> 
> Do you know what's wrong? And possibl another way to achieve what I want?

location /myapp/css/ {
    root  /home/stelios;
}

location /myapp/ {
    root  /home/stelios;
    try_files  $uri  $uri/  @php;
}

location @php {
    fastcgi_pass 127.0.0.1:9000;
    fastcgi_param SCRIPT_FILENAME /home/stelios/myapp/index.php;
    fastcgi_param QUERY_STRING    u=$uri;
    include fastcgi_params0;
}

fastcgi_params0 is a copy of fastcgi_params without QUERY_STRING parameter.


-- 
Igor Sysoev


From appa at perusio.net  Tue May 24 00:25:11 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Mon, 23 May 2011 21:25:11 +0100
Subject: another rewrite/try files issue
In-Reply-To: <BANLkTikxstF3+osH+bU-asLD28KzVP4BgA@mail.gmail.com>
References: <4DDA0454.5080101@gmail.com>
	<64a6135089b22eb9bfcfafb93ef9ceba.NginxMailingListEnglish@forum.nginx.org>
	<BANLkTikxstF3+osH+bU-asLD28KzVP4BgA@mail.gmail.com>
Message-ID: <87lixxkvvc.wl%appa@perusio.net>

On 23 Mai 2011 20h31 WEST, sgsfak at gmail.com wrote:

> 2011/5/23 brianmercer <nginx-forum at nginx.us>:
>> Oh, I see. ?Maybe this:
>>
>> location ~ /myapp/(?<myquery>.*) {
>> root /home/stelios;
>> try_files $uri /myapp/index.php?u=/$myquery;
>> }
>>
>> location = /myapp/index.php {
>> fastcgi_pass 127.0.0.1:9000;
>> fastcgi_param SCRIPT_FILENAME /home/stelios/myapp/index.php;
>> include fastcgi_params;
>> }
>
> Thanks again!
>
> I finally did the following which seems to work just fine:
>
> location ~ /myapp/(.*) {
> root /home/stelios/myapp;
> set $args u=/$1&$args;
> try_files /$1 @fallback;
> }

You can do it without any set directive. Try this.

location ~ /myapp/(?<myquery>.*)$ {
  root /home/stelios/myapp;
  try_files /$myquery /index.php?u=/$myquery&$args;
}

location = /index.php {
   root /home/stelios/myapp;     
   fastcgi_pass 127.0.0.1:9000;
   include fastcgi_params;
   ## Remove the line below if already defined in fastcgi_params.
   fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name; 
}

--- appa


From sbc19861004 at gmail.com  Tue May 24 06:49:45 2011
From: sbc19861004 at gmail.com (eagle sbc)
Date: Tue, 24 May 2011 10:49:45 +0800
Subject: SSI include infinite loop
Message-ID: <BANLkTi=u-pHdYdFA5P2=nAioCuyCija5Cw@mail.gmail.com>

Hello,

I've encountered a problem about "SSI include infinite loop".

We have a system that our editors can compose a html page from some
templates using SSI. Sometimes unwarily combination will cause infinite
loop, such as 'A include A', 'A include B, B include A', or even 'A include
B, B include C, C include A'. And this page will cause nginx eating up all
memories, and certainly won't serve any more unless we restart nginx.

I'm wondering is there any solution for this problem? Or other ways to avoid
this?

I've googled apache and lighttpd, and there also no official solution for
this, unless lightptd has a patch which will limit the whole 'include'
count, and I think it's just a tricky way to fix the infinite loop.

Thanks a lot for any help!


-- 
???????????????????
??????????????????????????
???????
??????????????
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110524/2164f89a/attachment-0001.html>

From mp3geek at gmail.com  Tue May 24 16:08:58 2011
From: mp3geek at gmail.com (Ryan B)
Date: Wed, 25 May 2011 00:08:58 +1200
Subject: Benchmarks of cherokee vs nginx
Message-ID: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>

This is static content (only http, and no php enabled), also using
gzip-static.in nginx (cache-io doesn't quite cut it in cherokee)

I found the cache too aggressive in cherokee, if I upload a newer file
I'd still keep serving the the cached file for a while (I wasn't
actually sure when it actually expired).. so I manually lowered the
expiry time for the cache (900secs), performance dives :/

Okay a quick break down of the stats..

Nginx-generated traffic is cut in half (thanks to gzip-static)  vs Cherokee
Nginx: 118Mb Ram, Cherokee: 260Mb
CPU: nginx is by-far using less, that race isn't even close.

A break down, http://i.imgur.com/JVO1w.png

Both web servers packages are compiled on ubuntu 11.04/openvz hosting.


From mdounin at mdounin.ru  Tue May 24 16:59:01 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 24 May 2011 16:59:01 +0400
Subject: SSI include infinite loop
In-Reply-To: <BANLkTi=u-pHdYdFA5P2=nAioCuyCija5Cw@mail.gmail.com>
References: <BANLkTi=u-pHdYdFA5P2=nAioCuyCija5Cw@mail.gmail.com>
Message-ID: <20110524125901.GW42265@mdounin.ru>

Hello!

On Tue, May 24, 2011 at 10:49:45AM +0800, eagle sbc wrote:

> Hello,
> 
> I've encountered a problem about "SSI include infinite loop".
> 
> We have a system that our editors can compose a html page from some
> templates using SSI. Sometimes unwarily combination will cause infinite
> loop, such as 'A include A', 'A include B, B include A', or even 'A include
> B, B include C, C include A'. And this page will cause nginx eating up all
> memories, and certainly won't serve any more unless we restart nginx.
> 
> I'm wondering is there any solution for this problem? Or other ways to avoid
> this?
> 
> I've googled apache and lighttpd, and there also no official solution for
> this, unless lightptd has a patch which will limit the whole 'include'
> count, and I think it's just a tricky way to fix the infinite loop.
> 
> Thanks a lot for any help!

Number of in-flight ssi includes used to be limited in nginx, 
though this limit was broken somewhere near 0.7.25.

This patch brings it back:

http://nginx.org/pipermail/nginx-devel/2011-February/000746.html

Maxim Dounin


From mp3geek at gmail.com  Tue May 24 17:02:36 2011
From: mp3geek at gmail.com (Ryan B)
Date: Wed, 25 May 2011 01:02:36 +1200
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
	<BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
	<BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>
Message-ID: <BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>

I'm guessing any form of vps setup's would have issues with
benchmarking.. I was hardly going to run this on my home connection or
pay for a uber dedi. I'm not trying to play favourites, I'm just
comparing the difference between the 2 web servers running on my vps,
and its pretty obvious looking a long/short-term graphs, using the "oh
you're on a openvz/vps, therefore this doesn't count" card doesn't
hold.

The graphs don't deviate much. The node isn't under any great load,
and it would show in the graphs (I could post longer 2-3-4 days but
they're pretty much the same).


2011/5/25 Tony Zakula <tonyzakula at gmail.com>:
>> Both web servers packages are compiled on ubuntu 11.04/openvz hosting.
>
> Using openvz hosting? ?That is not even a real test. ?There are so
> many variables, tweaks and adjustments when using OpenVZ that you
> could not possibly count on that bench mark even if you controlled the
> server. ?If you are using a hosting providers server, it makes it even
> more nebulous because they will be using their own controls. ?The
> memory and processing units are not even for real memory. ? With
> OpenVZ you can tweak those at will.
>
> The only way to get real benchmarks is on a real machine.
>
> Tony
>
>
> 2011/5/24 J?drzej Nowak <me at pigmej.eu>:
>> Hmm
>>
>> You can easy compress things in cherokee too. Just enable it in
>> cherokee-admin ;-)
>> For "gzip static" like behaviour you need to enable flcache (with
>> PURGE support for your config).
>>
>> IOCache in cherokee is caching only plain files when you don't use
>> gzip... That's why.
>>
>> Greetings,
>> J?drzej Nowak
>>
>>
>>
>> On Tue, May 24, 2011 at 2:08 PM, Ryan B <mp3geek at gmail.com> wrote:
>>> This is static content (only http, and no php enabled), also using
>>> gzip-static.in nginx (cache-io doesn't quite cut it in cherokee)
>>>
>>> I found the cache too aggressive in cherokee, if I upload a newer file
>>> I'd still keep serving the the cached file for a while (I wasn't
>>> actually sure when it actually expired).. so I manually lowered the
>>> expiry time for the cache (900secs), performance dives :/
>>>
>>> Okay a quick break down of the stats..
>>>
>>> Nginx-generated traffic is cut in half (thanks to gzip-static) ?vs Cherokee
>>> Nginx: 118Mb Ram, Cherokee: 260Mb
>>> CPU: nginx is by-far using less, that race isn't even close.
>>>
>>> A break down, http://i.imgur.com/JVO1w.png
>>>
>>> Both web servers packages are compiled on ubuntu 11.04/openvz hosting.
>>> _______________________________________________
>>> Cherokee mailing list
>>> Cherokee at lists.octality.com
>>> http://lists.octality.com/listinfo/cherokee
>>>
>> _______________________________________________
>> Cherokee mailing list
>> Cherokee at lists.octality.com
>> http://lists.octality.com/listinfo/cherokee
>>
>


From nginx-forum at nginx.us  Tue May 24 17:31:43 2011
From: nginx-forum at nginx.us (drmetal)
Date: Tue, 24 May 2011 09:31:43 -0400
Subject: cache large static files in memory
In-Reply-To: <5a11b2303ae5c2a8da3a2c89752907dc.NginxMailingList@forum.nginx.org>
References: <758641620906232200h4128ad0av103dbfafc78e612e@mail.gmail.com>
	<932ea6c90906232253t7d68e9f9n8267b56d9871fd63@mail.gmail.com>
	<758641620906232318o8fe6f40i9a21346aefc3d822@mail.gmail.com>
	<932ea6c90906240041o7ec3ad5dlb2c4c35ad6a15f3e@mail.gmail.com>
	<758641620906240354t1e0ed134xaf09987799fd77e0@forum.nginx.org>
	<5a11b2303ae5c2a8da3a2c89752907dc.NginxMailingList@forum.nginx.org>
Message-ID: <234e8b3486b4fd4cd7adacf815c70ea2.NginxMailingListEnglish@forum.nginx.org>

Hello, does this (linux RAM cache) apply to OpenVZ machines (Guests)?

Because when I run free -m command, i see 0 in buffers and cached.

Anyone knows how that goes in OpenVZ?

Thanks

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,3316,200352#msg-200352



From skebi69 at gmail.com  Tue May 24 18:00:49 2011
From: skebi69 at gmail.com (John Kennedy)
Date: Tue, 24 May 2011 15:00:49 +0100
Subject: Wildcard domains as vhost
Message-ID: <BANLkTiknoSccBzxb1YZX9bRtF2JiP5mvDw@mail.gmail.com>

I tried to google this but could not hit the right keywords (been a long
week)...

We have DNS set up so that all requests ending in .project.example.com (i.e.
bob.project.example.com) go to the same machine. How can I set up
sites-enabled so that all requests for *.project.example.com get treated the
same?

Thanks,
John

-- 
 John Kennedy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110524/9d75c399/attachment.html>

From mat999 at gmail.com  Tue May 24 18:03:57 2011
From: mat999 at gmail.com (SplitIce)
Date: Wed, 25 May 2011 00:03:57 +1000
Subject: Wildcard domains as vhost
In-Reply-To: <BANLkTiknoSccBzxb1YZX9bRtF2JiP5mvDw@mail.gmail.com>
References: <BANLkTiknoSccBzxb1YZX9bRtF2JiP5mvDw@mail.gmail.com>
Message-ID: <BANLkTi=PCtwMCShUSSpsj=bnVbbGt++=PA@mail.gmail.com>

look in the wiki at the server_name directive.  (wildcards)

On Wed, May 25, 2011 at 12:00 AM, John Kennedy <skebi69 at gmail.com> wrote:

> I tried to google this but could not hit the right keywords (been a long
> week)...
>
> We have DNS set up so that all requests ending in .project.example.com(i.e.
> bob.project.example.com) go to the same machine. How can I set up
> sites-enabled so that all requests for *.project.example.com get treated
> the same?
>
> Thanks,
> John
>
> --
>  John Kennedy
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110525/bdfc2329/attachment.html>

From skebi69 at gmail.com  Tue May 24 18:11:52 2011
From: skebi69 at gmail.com (John Kennedy)
Date: Tue, 24 May 2011 15:11:52 +0100
Subject: Wildcard domains as vhost
In-Reply-To: <BANLkTi=PCtwMCShUSSpsj=bnVbbGt++=PA@mail.gmail.com>
References: <BANLkTiknoSccBzxb1YZX9bRtF2JiP5mvDw@mail.gmail.com>
	<BANLkTi=PCtwMCShUSSpsj=bnVbbGt++=PA@mail.gmail.com>
Message-ID: <BANLkTik5dRx_UuDJWSTD5QSkqnsMRKgZ=w@mail.gmail.com>

That was what I was looking for.
Thanks.
John

On Tue, May 24, 2011 at 15:03, SplitIce <mat999 at gmail.com> wrote:

> look in the wiki at the server_name directive.  (wildcards)
>
> On Wed, May 25, 2011 at 12:00 AM, John Kennedy <skebi69 at gmail.com> wrote:
>
>> I tried to google this but could not hit the right keywords (been a long
>> week)...
>>
>> We have DNS set up so that all requests ending in .project.example.com(i.e.
>> bob.project.example.com) go to the same machine. How can I set up
>> sites-enabled so that all requests for *.project.example.com get treated
>> the same?
>>
>> Thanks,
>> John
>>
>> --
>>  John Kennedy
>>
>>
>> _______________________________________________
>> nginx mailing list
>> nginx at nginx.org
>> http://nginx.org/mailman/listinfo/nginx
>>
>>
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>


-- 
 John Kennedy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110524/6c9c53ff/attachment.html>

From appa at perusio.net  Tue May 24 18:13:02 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Tue, 24 May 2011 15:13:02 +0100
Subject: Wildcard domains as vhost
In-Reply-To: <BANLkTiknoSccBzxb1YZX9bRtF2JiP5mvDw@mail.gmail.com>
References: <BANLkTiknoSccBzxb1YZX9bRtF2JiP5mvDw@mail.gmail.com>
Message-ID: <87ipt0kx01.wl%appa@perusio.net>

On 24 Mai 2011 15h00 WEST, skebi69 at gmail.com wrote:

> [1 <multipart/alternative (7bit)>] [1.1 <text/plain; ISO-8859-1
> (7bit)>] I tried to google this but could not hit the right keywords
> (been a long week)...
>
> We have DNS set up so that all requests ending in
> .project.example.com (i.e.  bob.project.example.com) go to the same
> machine. How can I set up sites-enabled so that all requests for
> *.project.example.com get treated the same?

Cf. http://nginx.org/en/docs/http/server_names.html#wildcard_names

server {
  server_name *.project.example.com;
  (...)
}

--- appa



From cliff at develix.com  Tue May 24 18:14:25 2011
From: cliff at develix.com (Cliff Wells)
Date: Tue, 24 May 2011 07:14:25 -0700
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
	<BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
	<BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>
	<BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>
Message-ID: <1306246465.2327.20.camel@portable-evil>

On Wed, 2011-05-25 at 01:02 +1200, Ryan B wrote:
> I'm guessing any form of vps setup's would have issues with
> benchmarking.. I was hardly going to run this on my home connection or
> pay for a uber dedi.

Actually, it's a fine benchmark so long as both HTTP servers are running
in the same environment.  Can you extrapolate and draw conclusions about
how they might run on different hardware or in a different environment?
Maybe with a grain of salt, but that's the case with any benchmark, so
complaining about OpenVZ tuning is pointless.

Putting it on dedicated hardware isn't terribly useful either, since
fewer people do that now. Even dedicated servers are virtualized for
management reasons.

The only issue I'd have is if the node has other VE's on it that you
don't have control over since those could clearly affect the outcome.
Still I'd expect deviations to show from multiple runs that would reveal
that type of issue.

Cliff




From cliff at develix.com  Tue May 24 18:39:39 2011
From: cliff at develix.com (Cliff Wells)
Date: Tue, 24 May 2011 07:39:39 -0700
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <BANLkTi=HhKgmNKxKQPTNnJ7R+CkwvbUtcw@mail.gmail.com>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
	<BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
	<BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>
	<BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>
	<1306246465.2327.20.camel@portable-evil>
	<BANLkTi=HhKgmNKxKQPTNnJ7R+CkwvbUtcw@mail.gmail.com>
Message-ID: <1306247979.2327.31.camel@portable-evil>

On Tue, 2011-05-24 at 09:25 -0500, Tony Zakula wrote:
> I am not complaining about OpenVZ.  I deploy servers on a regular
> basis using OpenVZ.  I also deploy using KVM, and others.  There is a
> massive difference between OpenVZ and KVM in implementation, system
> stability, etc.  They are entirely different animals.  You cannot
> discount the hypervisor.  The hypervisor can make or break your
> application depending on how you're application is structured.  For
> instance, Java apps run great on KVM hypervisors, but really poor if
> at all on OpenVZ unless you tune the OpenVZ instance to meet your Java
> apps needs.  The reason for that is the way Java handles memory.
> Threads is another big issue hypervisors. To say that the hypervisor
> has no effect on application performance is not accurate at all.

So are you suggesting no benchmark is ever valid unless it is run on
every hypervisor available as well as on the bare metal?  Clearly every
benchmark has defined parameters, and in this case OpenVZ was one of
those parameters.

So long as the benchmark is defined as "Nginx vs Cherokee under OpenVZ"
then it's a perfectly valid benchmark.

Cliff



From cliff at develix.com  Tue May 24 18:48:19 2011
From: cliff at develix.com (Cliff Wells)
Date: Tue, 24 May 2011 07:48:19 -0700
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <1306247979.2327.31.camel@portable-evil>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
	<BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
	<BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>
	<BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>
	<1306246465.2327.20.camel@portable-evil>
	<BANLkTi=HhKgmNKxKQPTNnJ7R+CkwvbUtcw@mail.gmail.com>
	<1306247979.2327.31.camel@portable-evil>
Message-ID: <1306248499.2327.34.camel@portable-evil>

On Tue, 2011-05-24 at 07:39 -0700, Cliff Wells wrote:
> On Tue, 2011-05-24 at 09:25 -0500, Tony Zakula wrote:
> > I am not complaining about OpenVZ.  I deploy servers on a regular
> > basis using OpenVZ.  I also deploy using KVM, and others.  There is a
> > massive difference between OpenVZ and KVM in implementation, system
> > stability, etc.  They are entirely different animals.  You cannot
> > discount the hypervisor.  The hypervisor can make or break your
> > application depending on how you're application is structured.  For
> > instance, Java apps run great on KVM hypervisors, but really poor if
> > at all on OpenVZ unless you tune the OpenVZ instance to meet your Java
> > apps needs.  The reason for that is the way Java handles memory.
> > Threads is another big issue hypervisors. To say that the hypervisor
> > has no effect on application performance is not accurate at all.
> 
> So are you suggesting no benchmark is ever valid unless it is run on
> every hypervisor available as well as on the bare metal?  Clearly every
> benchmark has defined parameters, and in this case OpenVZ was one of
> those parameters.
> 
> So long as the benchmark is defined as "Nginx vs Cherokee under OpenVZ"
> then it's a perfectly valid benchmark.

Let me clarify that last sentence: there are other issues that might
invalidate this benchmark (should be run on a dedicated LAN and on a
dedicated node at least), but the virtualization (or lack of) isn't one
of those factors.

Cliff



From cliff at develix.com  Tue May 24 19:04:38 2011
From: cliff at develix.com (Cliff Wells)
Date: Tue, 24 May 2011 08:04:38 -0700
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <BANLkTikT3N6n7tQg8Oso-nPs3pmkFt0stA@mail.gmail.com>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
	<BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
	<BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>
	<BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>
	<1306246465.2327.20.camel@portable-evil>
	<BANLkTi=HhKgmNKxKQPTNnJ7R+CkwvbUtcw@mail.gmail.com>
	<1306247979.2327.31.camel@portable-evil>
	<BANLkTikT3N6n7tQg8Oso-nPs3pmkFt0stA@mail.gmail.com>
Message-ID: <1306249478.2327.37.camel@portable-evil>

On Tue, 2011-05-24 at 09:50 -0500, Tony Zakula wrote:
> Was there any info on how OpenVZ was setup or how many other
> containers were running on the machine, Do we know what the load was
> on the machine at each stage of the tests?  We really do not know
> anything.  The web servers were not even setup the same.  The hosting
> outfit could see a system spike and automatically kick in a limit that
> we do not know about. I do that all the time. That is a common
> practice.  If you do not control the system, you cannot do a valid
> benchmark.

I agree with that.  I only disagree with your earlier, broader
assessment.

Cliff




From jelledejong at powercraft.nl  Tue May 24 19:19:10 2011
From: jelledejong at powercraft.nl (Jelle de Jong)
Date: Tue, 24 May 2011 17:19:10 +0200
Subject: TCP: time wait bucket table overflow
Message-ID: <4DDBCC6E.7080806@powercraft.nl>

I am just going to pop-up this question. My nginx was doing about 10%
cpu load and I was getting:

May 24 15:58:43 yasmin kernel: [  983.444403] TCP: time wait bucket
table overflow

I know I can fine-tune my /proc/sys/net/ipv4/tcp_max_tw_buckets and some
other stuff. But I am pretty sure somebody here has more experience with
tcp tuning.

What are good Linux kernel settings for a proxy_pass / load balancer
nginx server?

Some URL's to background articles would be nice.

In my configuration it passes everything to Apache servers as back-ends.

Thanks in advance,

Kind regards,

Jelle de Jong


From me at pigmej.eu  Tue May 24 16:17:07 2011
From: me at pigmej.eu (=?UTF-8?Q?J=C4=99drzej_Nowak?=)
Date: Tue, 24 May 2011 14:17:07 +0200
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
Message-ID: <BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>

Hmm

You can easy compress things in cherokee too. Just enable it in
cherokee-admin ;-)
For "gzip static" like behaviour you need to enable flcache (with
PURGE support for your config).

IOCache in cherokee is caching only plain files when you don't use
gzip... That's why.

Greetings,
J?drzej Nowak



On Tue, May 24, 2011 at 2:08 PM, Ryan B <mp3geek at gmail.com> wrote:
> This is static content (only http, and no php enabled), also using
> gzip-static.in nginx (cache-io doesn't quite cut it in cherokee)
>
> I found the cache too aggressive in cherokee, if I upload a newer file
> I'd still keep serving the the cached file for a while (I wasn't
> actually sure when it actually expired).. so I manually lowered the
> expiry time for the cache (900secs), performance dives :/
>
> Okay a quick break down of the stats..
>
> Nginx-generated traffic is cut in half (thanks to gzip-static) ?vs Cherokee
> Nginx: 118Mb Ram, Cherokee: 260Mb
> CPU: nginx is by-far using less, that race isn't even close.
>
> A break down, http://i.imgur.com/JVO1w.png
>
> Both web servers packages are compiled on ubuntu 11.04/openvz hosting.
> _______________________________________________
> Cherokee mailing list
> Cherokee at lists.octality.com
> http://lists.octality.com/listinfo/cherokee
>



From tonyzakula at gmail.com  Tue May 24 16:33:11 2011
From: tonyzakula at gmail.com (Tony Zakula)
Date: Tue, 24 May 2011 07:33:11 -0500
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
	<BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
Message-ID: <BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>

> Both web servers packages are compiled on ubuntu 11.04/openvz hosting.

Using openvz hosting?  That is not even a real test.  There are so
many variables, tweaks and adjustments when using OpenVZ that you
could not possibly count on that bench mark even if you controlled the
server.  If you are using a hosting providers server, it makes it even
more nebulous because they will be using their own controls.  The
memory and processing units are not even for real memory.   With
OpenVZ you can tweak those at will.

The only way to get real benchmarks is on a real machine.

Tony


2011/5/24 J?drzej Nowak <me at pigmej.eu>:
> Hmm
>
> You can easy compress things in cherokee too. Just enable it in
> cherokee-admin ;-)
> For "gzip static" like behaviour you need to enable flcache (with
> PURGE support for your config).
>
> IOCache in cherokee is caching only plain files when you don't use
> gzip... That's why.
>
> Greetings,
> J?drzej Nowak
>
>
>
> On Tue, May 24, 2011 at 2:08 PM, Ryan B <mp3geek at gmail.com> wrote:
>> This is static content (only http, and no php enabled), also using
>> gzip-static.in nginx (cache-io doesn't quite cut it in cherokee)
>>
>> I found the cache too aggressive in cherokee, if I upload a newer file
>> I'd still keep serving the the cached file for a while (I wasn't
>> actually sure when it actually expired).. so I manually lowered the
>> expiry time for the cache (900secs), performance dives :/
>>
>> Okay a quick break down of the stats..
>>
>> Nginx-generated traffic is cut in half (thanks to gzip-static) ?vs Cherokee
>> Nginx: 118Mb Ram, Cherokee: 260Mb
>> CPU: nginx is by-far using less, that race isn't even close.
>>
>> A break down, http://i.imgur.com/JVO1w.png
>>
>> Both web servers packages are compiled on ubuntu 11.04/openvz hosting.
>> _______________________________________________
>> Cherokee mailing list
>> Cherokee at lists.octality.com
>> http://lists.octality.com/listinfo/cherokee
>>
> _______________________________________________
> Cherokee mailing list
> Cherokee at lists.octality.com
> http://lists.octality.com/listinfo/cherokee
>



From me at pigmej.eu  Tue May 24 17:04:44 2011
From: me at pigmej.eu (=?UTF-8?Q?J=C4=99drzej_Nowak?=)
Date: Tue, 24 May 2011 15:04:44 +0200
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
	<BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
	<BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>
	<BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>
Message-ID: <BANLkTikzV=EVZ1ZPjiFUMTqi=+-qYyV=pg@mail.gmail.com>

Well...

You compared cherokee without GZIP to nginx with GZIP... And you're
comparing a memory used by cache...
It's a miss point.

Greetings,
J?drzej Nowak



On Tue, May 24, 2011 at 3:02 PM, Ryan B <mp3geek at gmail.com> wrote:
> I'm guessing any form of vps setup's would have issues with
> benchmarking.. I was hardly going to run this on my home connection or
> pay for a uber dedi. I'm not trying to play favourites, I'm just
> comparing the difference between the 2 web servers running on my vps,
> and its pretty obvious looking a long/short-term graphs, using the "oh
> you're on a openvz/vps, therefore this doesn't count" card doesn't
> hold.
>
> The graphs don't deviate much. The node isn't under any great load,
> and it would show in the graphs (I could post longer 2-3-4 days but
> they're pretty much the same).
>
>
> 2011/5/25 Tony Zakula <tonyzakula at gmail.com>:
>>> Both web servers packages are compiled on ubuntu 11.04/openvz hosting.
>>
>> Using openvz hosting? ?That is not even a real test. ?There are so
>> many variables, tweaks and adjustments when using OpenVZ that you
>> could not possibly count on that bench mark even if you controlled the
>> server. ?If you are using a hosting providers server, it makes it even
>> more nebulous because they will be using their own controls. ?The
>> memory and processing units are not even for real memory. ? With
>> OpenVZ you can tweak those at will.
>>
>> The only way to get real benchmarks is on a real machine.
>>
>> Tony
>>
>>
>> 2011/5/24 J?drzej Nowak <me at pigmej.eu>:
>>> Hmm
>>>
>>> You can easy compress things in cherokee too. Just enable it in
>>> cherokee-admin ;-)
>>> For "gzip static" like behaviour you need to enable flcache (with
>>> PURGE support for your config).
>>>
>>> IOCache in cherokee is caching only plain files when you don't use
>>> gzip... That's why.
>>>
>>> Greetings,
>>> J?drzej Nowak
>>>
>>>
>>>
>>> On Tue, May 24, 2011 at 2:08 PM, Ryan B <mp3geek at gmail.com> wrote:
>>>> This is static content (only http, and no php enabled), also using
>>>> gzip-static.in nginx (cache-io doesn't quite cut it in cherokee)
>>>>
>>>> I found the cache too aggressive in cherokee, if I upload a newer file
>>>> I'd still keep serving the the cached file for a while (I wasn't
>>>> actually sure when it actually expired).. so I manually lowered the
>>>> expiry time for the cache (900secs), performance dives :/
>>>>
>>>> Okay a quick break down of the stats..
>>>>
>>>> Nginx-generated traffic is cut in half (thanks to gzip-static) ?vs Cherokee
>>>> Nginx: 118Mb Ram, Cherokee: 260Mb
>>>> CPU: nginx is by-far using less, that race isn't even close.
>>>>
>>>> A break down, http://i.imgur.com/JVO1w.png
>>>>
>>>> Both web servers packages are compiled on ubuntu 11.04/openvz hosting.
>>>> _______________________________________________
>>>> Cherokee mailing list
>>>> Cherokee at lists.octality.com
>>>> http://lists.octality.com/listinfo/cherokee
>>>>
>>> _______________________________________________
>>> Cherokee mailing list
>>> Cherokee at lists.octality.com
>>> http://lists.octality.com/listinfo/cherokee
>>>
>>
> _______________________________________________
> Cherokee mailing list
> Cherokee at lists.octality.com
> http://lists.octality.com/listinfo/cherokee
>



From tonyzakula at gmail.com  Tue May 24 17:15:12 2011
From: tonyzakula at gmail.com (Tony Zakula)
Date: Tue, 24 May 2011 08:15:12 -0500
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <BANLkTikzV=EVZ1ZPjiFUMTqi=+-qYyV=pg@mail.gmail.com>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
	<BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
	<BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>
	<BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>
	<BANLkTikzV=EVZ1ZPjiFUMTqi=+-qYyV=pg@mail.gmail.com>
Message-ID: <BANLkTimaDRkKjd87UubfTgSRoMoCn4Sutw@mail.gmail.com>

Right, as with any benchmark, it is better to say up front what
conditions you tested with, how you tested, and what variables were
involved.

So myhost.com, with this VPS package, with Cherokee configured this
way, and Nginx configured this way, etc.

It is not a "this doesn't count card" to you, but it is for most
people.  I am happy to see people share, after all, the conclusion
would be, migrate to a host that better supports Cherokee if you love
using it, or ask your host to tweak their VPS if you cannot get the
benchmarks acceptable.  The conclusion is not that Nginx is better
than Cherokee.

Even those memory and cpu numbers are totally fake.  They are software
numbers fed by OpenVZ to the container.

Tony


2011/5/24 J?drzej Nowak <me at pigmej.eu>:
> Well...
>
> You compared cherokee without GZIP to nginx with GZIP... And you're
> comparing a memory used by cache...
> It's a miss point.
>
> Greetings,
> J?drzej Nowak
>
>
>
> On Tue, May 24, 2011 at 3:02 PM, Ryan B <mp3geek at gmail.com> wrote:
>> I'm guessing any form of vps setup's would have issues with
>> benchmarking.. I was hardly going to run this on my home connection or
>> pay for a uber dedi. I'm not trying to play favourites, I'm just
>> comparing the difference between the 2 web servers running on my vps,
>> and its pretty obvious looking a long/short-term graphs, using the "oh
>> you're on a openvz/vps, therefore this doesn't count" card doesn't
>> hold.
>>
>> The graphs don't deviate much. The node isn't under any great load,
>> and it would show in the graphs (I could post longer 2-3-4 days but
>> they're pretty much the same).
>>
>>
>> 2011/5/25 Tony Zakula <tonyzakula at gmail.com>:
>>>> Both web servers packages are compiled on ubuntu 11.04/openvz hosting.
>>>
>>> Using openvz hosting? ?That is not even a real test. ?There are so
>>> many variables, tweaks and adjustments when using OpenVZ that you
>>> could not possibly count on that bench mark even if you controlled the
>>> server. ?If you are using a hosting providers server, it makes it even
>>> more nebulous because they will be using their own controls. ?The
>>> memory and processing units are not even for real memory. ? With
>>> OpenVZ you can tweak those at will.
>>>
>>> The only way to get real benchmarks is on a real machine.
>>>
>>> Tony
>>>
>>>
>>> 2011/5/24 J?drzej Nowak <me at pigmej.eu>:
>>>> Hmm
>>>>
>>>> You can easy compress things in cherokee too. Just enable it in
>>>> cherokee-admin ;-)
>>>> For "gzip static" like behaviour you need to enable flcache (with
>>>> PURGE support for your config).
>>>>
>>>> IOCache in cherokee is caching only plain files when you don't use
>>>> gzip... That's why.
>>>>
>>>> Greetings,
>>>> J?drzej Nowak
>>>>
>>>>
>>>>
>>>> On Tue, May 24, 2011 at 2:08 PM, Ryan B <mp3geek at gmail.com> wrote:
>>>>> This is static content (only http, and no php enabled), also using
>>>>> gzip-static.in nginx (cache-io doesn't quite cut it in cherokee)
>>>>>
>>>>> I found the cache too aggressive in cherokee, if I upload a newer file
>>>>> I'd still keep serving the the cached file for a while (I wasn't
>>>>> actually sure when it actually expired).. so I manually lowered the
>>>>> expiry time for the cache (900secs), performance dives :/
>>>>>
>>>>> Okay a quick break down of the stats..
>>>>>
>>>>> Nginx-generated traffic is cut in half (thanks to gzip-static) ?vs Cherokee
>>>>> Nginx: 118Mb Ram, Cherokee: 260Mb
>>>>> CPU: nginx is by-far using less, that race isn't even close.
>>>>>
>>>>> A break down, http://i.imgur.com/JVO1w.png
>>>>>
>>>>> Both web servers packages are compiled on ubuntu 11.04/openvz hosting.
>>>>> _______________________________________________
>>>>> Cherokee mailing list
>>>>> Cherokee at lists.octality.com
>>>>> http://lists.octality.com/listinfo/cherokee
>>>>>
>>>> _______________________________________________
>>>> Cherokee mailing list
>>>> Cherokee at lists.octality.com
>>>> http://lists.octality.com/listinfo/cherokee
>>>>
>>>
>> _______________________________________________
>> Cherokee mailing list
>> Cherokee at lists.octality.com
>> http://lists.octality.com/listinfo/cherokee
>>
>



From tonyzakula at gmail.com  Tue May 24 18:25:19 2011
From: tonyzakula at gmail.com (Tony Zakula)
Date: Tue, 24 May 2011 09:25:19 -0500
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <1306246465.2327.20.camel@portable-evil>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
	<BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
	<BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>
	<BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>
	<1306246465.2327.20.camel@portable-evil>
Message-ID: <BANLkTi=HhKgmNKxKQPTNnJ7R+CkwvbUtcw@mail.gmail.com>

I am not complaining about OpenVZ.  I deploy servers on a regular
basis using OpenVZ.  I also deploy using KVM, and others.  There is a
massive difference between OpenVZ and KVM in implementation, system
stability, etc.  They are entirely different animals.  You cannot
discount the hypervisor.  The hypervisor can make or break your
application depending on how you're application is structured.  For
instance, Java apps run great on KVM hypervisors, but really poor if
at all on OpenVZ unless you tune the OpenVZ instance to meet your Java
apps needs.  The reason for that is the way Java handles memory.
Threads is another big issue hypervisors. To say that the hypervisor
has no effect on application performance is not accurate at all.

Tony


On Tue, May 24, 2011 at 9:14 AM, Cliff Wells <cliff at develix.com> wrote:
> On Wed, 2011-05-25 at 01:02 +1200, Ryan B wrote:
>> I'm guessing any form of vps setup's would have issues with
>> benchmarking.. I was hardly going to run this on my home connection or
>> pay for a uber dedi.
>
> Actually, it's a fine benchmark so long as both HTTP servers are running
> in the same environment. ?Can you extrapolate and draw conclusions about
> how they might run on different hardware or in a different environment?
> Maybe with a grain of salt, but that's the case with any benchmark, so
> complaining about OpenVZ tuning is pointless.
>
> Putting it on dedicated hardware isn't terribly useful either, since
> fewer people do that now. Even dedicated servers are virtualized for
> management reasons.
>
> The only issue I'd have is if the node has other VE's on it that you
> don't have control over since those could clearly affect the outcome.
> Still I'd expect deviations to show from multiple runs that would reveal
> that type of issue.
>
> Cliff
>
>
>



From tonyzakula at gmail.com  Tue May 24 18:50:52 2011
From: tonyzakula at gmail.com (Tony Zakula)
Date: Tue, 24 May 2011 09:50:52 -0500
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <1306247979.2327.31.camel@portable-evil>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
	<BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
	<BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>
	<BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>
	<1306246465.2327.20.camel@portable-evil>
	<BANLkTi=HhKgmNKxKQPTNnJ7R+CkwvbUtcw@mail.gmail.com>
	<1306247979.2327.31.camel@portable-evil>
Message-ID: <BANLkTikT3N6n7tQg8Oso-nPs3pmkFt0stA@mail.gmail.com>

Was there any info on how OpenVZ was setup or how many other
containers were running on the machine, Do we know what the load was
on the machine at each stage of the tests?  We really do not know
anything.  The web servers were not even setup the same.  The hosting
outfit could see a system spike and automatically kick in a limit that
we do not know about. I do that all the time. That is a common
practice.  If you do not control the system, you cannot do a valid
benchmark.

On Tue, May 24, 2011 at 9:39 AM, Cliff Wells <cliff at develix.com> wrote:
> On Tue, 2011-05-24 at 09:25 -0500, Tony Zakula wrote:
>> I am not complaining about OpenVZ. ?I deploy servers on a regular
>> basis using OpenVZ. ?I also deploy using KVM, and others. ?There is a
>> massive difference between OpenVZ and KVM in implementation, system
>> stability, etc. ?They are entirely different animals. ?You cannot
>> discount the hypervisor. ?The hypervisor can make or break your
>> application depending on how you're application is structured. ?For
>> instance, Java apps run great on KVM hypervisors, but really poor if
>> at all on OpenVZ unless you tune the OpenVZ instance to meet your Java
>> apps needs. ?The reason for that is the way Java handles memory.
>> Threads is another big issue hypervisors. To say that the hypervisor
>> has no effect on application performance is not accurate at all.
>
> So are you suggesting no benchmark is ever valid unless it is run on
> every hypervisor available as well as on the bare metal? ?Clearly every
> benchmark has defined parameters, and in this case OpenVZ was one of
> those parameters.
>
> So long as the benchmark is defined as "Nginx vs Cherokee under OpenVZ"
> then it's a perfectly valid benchmark.
>
> Cliff
>
>



From tonyzakula at gmail.com  Tue May 24 18:52:41 2011
From: tonyzakula at gmail.com (Tony Zakula)
Date: Tue, 24 May 2011 09:52:41 -0500
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <BANLkTikT3N6n7tQg8Oso-nPs3pmkFt0stA@mail.gmail.com>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
	<BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
	<BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>
	<BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>
	<1306246465.2327.20.camel@portable-evil>
	<BANLkTi=HhKgmNKxKQPTNnJ7R+CkwvbUtcw@mail.gmail.com>
	<1306247979.2327.31.camel@portable-evil>
	<BANLkTikT3N6n7tQg8Oso-nPs3pmkFt0stA@mail.gmail.com>
Message-ID: <BANLkTinKHeTmzrmN+6gEEEQkF-snTHPc6A@mail.gmail.com>

And you certainly cannot report on memory consumption relying on
OpenVZ fake memory numbers.

On Tue, May 24, 2011 at 9:50 AM, Tony Zakula <tonyzakula at gmail.com> wrote:
> Was there any info on how OpenVZ was setup or how many other
> containers were running on the machine, Do we know what the load was
> on the machine at each stage of the tests? ?We really do not know
> anything. ?The web servers were not even setup the same. ?The hosting
> outfit could see a system spike and automatically kick in a limit that
> we do not know about. I do that all the time. That is a common
> practice. ?If you do not control the system, you cannot do a valid
> benchmark.
>
> On Tue, May 24, 2011 at 9:39 AM, Cliff Wells <cliff at develix.com> wrote:
>> On Tue, 2011-05-24 at 09:25 -0500, Tony Zakula wrote:
>>> I am not complaining about OpenVZ. ?I deploy servers on a regular
>>> basis using OpenVZ. ?I also deploy using KVM, and others. ?There is a
>>> massive difference between OpenVZ and KVM in implementation, system
>>> stability, etc. ?They are entirely different animals. ?You cannot
>>> discount the hypervisor. ?The hypervisor can make or break your
>>> application depending on how you're application is structured. ?For
>>> instance, Java apps run great on KVM hypervisors, but really poor if
>>> at all on OpenVZ unless you tune the OpenVZ instance to meet your Java
>>> apps needs. ?The reason for that is the way Java handles memory.
>>> Threads is another big issue hypervisors. To say that the hypervisor
>>> has no effect on application performance is not accurate at all.
>>
>> So are you suggesting no benchmark is ever valid unless it is run on
>> every hypervisor available as well as on the bare metal? ?Clearly every
>> benchmark has defined parameters, and in this case OpenVZ was one of
>> those parameters.
>>
>> So long as the benchmark is defined as "Nginx vs Cherokee under OpenVZ"
>> then it's a perfectly valid benchmark.
>>
>> Cliff
>>
>>
>



From tonyzakula at gmail.com  Tue May 24 19:12:22 2011
From: tonyzakula at gmail.com (Tony Zakula)
Date: Tue, 24 May 2011 10:12:22 -0500
Subject: [Cherokee] Benchmarks of cherokee vs nginx
In-Reply-To: <1306249478.2327.37.camel@portable-evil>
References: <BANLkTi=4JYHoLJbA3HcOQeZaJH=VAj=QHA@mail.gmail.com>
	<BANLkTimUzm8fkBOk-hBz32rUFPyMxOqXKw@mail.gmail.com>
	<BANLkTi=ksVFTxGS4Q7vUW2Hfep1+q9b0gQ@mail.gmail.com>
	<BANLkTimkRRoaxWfineoQGwYG2edQTX2NNw@mail.gmail.com>
	<1306246465.2327.20.camel@portable-evil>
	<BANLkTi=HhKgmNKxKQPTNnJ7R+CkwvbUtcw@mail.gmail.com>
	<1306247979.2327.31.camel@portable-evil>
	<BANLkTikT3N6n7tQg8Oso-nPs3pmkFt0stA@mail.gmail.com>
	<1306249478.2327.37.camel@portable-evil>
Message-ID: <BANLkTimAw1bZ9krKw7K7Ng3F0KBOK3n6FA@mail.gmail.com>

On Tue, May 24, 2011 at 10:04 AM, Cliff Wells <cliff at develix.com> wrote:
> On Tue, 2011-05-24 at 09:50 -0500, Tony Zakula wrote:
>> Was there any info on how OpenVZ was setup or how many other
>> containers were running on the machine, Do we know what the load was
>> on the machine at each stage of the tests? ?We really do not know
>> anything. ?The web servers were not even setup the same. ?The hosting
>> outfit could see a system spike and automatically kick in a limit that
>> we do not know about. I do that all the time. That is a common
>> practice. ?If you do not control the system, you cannot do a valid
>> benchmark.
>
> I agree with that. ?I only disagree with your earlier, broader
> assessment.
>
> Cliff

I should have been more clear.  Point well taken.



From lpr at dyb.fm  Tue May 24 20:53:29 2011
From: lpr at dyb.fm (Felipe aka lpr)
Date: Tue, 24 May 2011 13:53:29 -0300
Subject: cache large static files in memory
In-Reply-To: <234e8b3486b4fd4cd7adacf815c70ea2.NginxMailingListEnglish@forum.nginx.org>
References: <758641620906232200h4128ad0av103dbfafc78e612e@mail.gmail.com>
	<932ea6c90906232253t7d68e9f9n8267b56d9871fd63@mail.gmail.com>
	<758641620906232318o8fe6f40i9a21346aefc3d822@mail.gmail.com>
	<932ea6c90906240041o7ec3ad5dlb2c4c35ad6a15f3e@mail.gmail.com>
	<758641620906240354t1e0ed134xaf09987799fd77e0@forum.nginx.org>
	<5a11b2303ae5c2a8da3a2c89752907dc.NginxMailingList@forum.nginx.org>
	<234e8b3486b4fd4cd7adacf815c70ea2.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BAAAAD4A-8F20-4F60-B97B-A1298E79D124@dyb.fm>

openvz uses conteiner virtualization.

The memory is shared with other vm's.

The configuration of yourvz.conf show how much is your guaranteed memory, and bursted memory.

And its not possible add a swap disk using swapon to a file image.

Em 24/05/2011, ?s 10:31, drmetal escreveu:

> Hello, does this (linux RAM cache) apply to OpenVZ machines (Guests)?
> 
> Because when I run free -m command, i see 0 in buffers and cached.
> 
> Anyone knows how that goes in OpenVZ?
> 
> Thanks
> 
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,3316,200352#msg-200352
> 
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx



From nginx-forum at nginx.us  Tue May 24 23:25:25 2011
From: nginx-forum at nginx.us (Trazzt)
Date: Tue, 24 May 2011 15:25:25 -0400
Subject: Configuring phpmyadmin
Message-ID: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>

Hi,

I recently installed Debian 6.0.1a Squeeze with Nginx and phpmyadmin
When I go to 'http://localhost' phpmyadmin pops up.
What I want is that phpmyadmin is located at
'http://localhost/phpmyadmin/'

I already tried to add this in nginx.conf. It didn't work

location /phpmyadmin {
alias /usr/share/phpmyadmin/;
}

This is what my nginx.conf looks like:

_____________________________________
user www-data;
worker_processes  1;

error_log  /var/log/nginx/error.log;
pid        /var/run/nginx.pid;

events {
    worker_connections  1024;
    # multi_accept on;
}

http {
    include       /etc/nginx/mime.types;

    access_log  /var/log/nginx/access.log;

    sendfile        on;
    #tcp_nopush     on;

    #keepalive_timeout  0;
    keepalive_timeout  65;
    tcp_nodelay        on;

    gzip  on;
    gzip_disable "MSIE [1-6]\.(?!.*SV1)";

    include /etc/nginx/conf.d/*.conf;
    include /etc/nginx/sites-enabled/*;
}
_____________________________________

Can anybody help me configuring this the right way?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200488,200488#msg-200488



From nginx-forum at nginx.us  Wed May 25 01:04:11 2011
From: nginx-forum at nginx.us (urschrei)
Date: Tue, 24 May 2011 17:04:11 -0400
Subject: Provide site-specific SSL cert on behalf of clients
Message-ID: <bb893cda9c86fc398927edad02f87284.NginxMailingListEnglish@forum.nginx.org>

I'm wondering if I can configure nginx to provide a PEM SSL cert for a
specific site on behalf of clients on my subnet, and how I might go
about configuring, testing this. I have a cert in PEM format, and nginx
0.8.54 with SSL support, running on Ubuntu 11.04.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200521,200521#msg-200521



From nginx-forum at nginx.us  Wed May 25 01:14:02 2011
From: nginx-forum at nginx.us (drmetal)
Date: Tue, 24 May 2011 17:14:02 -0400
Subject: cache large static files in memory
In-Reply-To: <BAAAAD4A-8F20-4F60-B97B-A1298E79D124@dyb.fm>
References: <BAAAAD4A-8F20-4F60-B97B-A1298E79D124@dyb.fm>
Message-ID: <adf800ea887c20f32f8ca29c2ccafe38.NginxMailingListEnglish@forum.nginx.org>

Thanks for the answer but this doesn't answer my question. Obviously I
have shared RAM. My question was, does linux store frequently used files
in RAM ? Because this is usually reported in the free -m command
output.

But in the case of OpenVZ VPS I see nothing (0) in the cached and
buffers column.

I researched a little, the system caches files, they are just not
reported.

Felipe aka lpr Wrote:
-------------------------------------------------------
> openvz uses conteiner virtualization.
> 
> The memory is shared with other vm's.
> 
> The configuration of yourvz.conf show how much is
> your guaranteed memory, and bursted memory.
> 
> And its not possible add a swap disk using swapon
> to a file image.
> 
> Em 24/05/2011, ?s 10:31, drmetal escreveu:
> 
> > Hello, does this (linux RAM cache) apply to
> OpenVZ machines (Guests)?
> > 
> > Because when I run free -m command, i see 0 in
> buffers and cached.
> > 
> > Anyone knows how that goes in OpenVZ?
> > 
> > Thanks
> > 
> > Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,3316,200352#msg-
> 200352
> > 
> > 
> > _______________________________________________
> > nginx mailing list
> > nginx at nginx.org
> > http://nginx.org/mailman/listinfo/nginx
> 
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,3316,200528#msg-200528



From caldcv at gmail.com  Wed May 25 03:15:26 2011
From: caldcv at gmail.com (Chris)
Date: Tue, 24 May 2011 19:15:26 -0400
Subject: Configuring phpmyadmin
In-Reply-To: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
References: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTi=_Cr46Dktt6kuuT=it6fMKBAFkYg@mail.gmail.com>

Post your /etc/nginx/sites-enabled/ files and also, keeping phpmyadmin
on a public directory is bad since bots / scripts automatically scan
for that in URLs. Restrict it with .htaccess if you want to keep
phpmyadmin as the directory name or change it to something like
"fuzzykittens" to hide it.

On 5/24/11, Trazzt <nginx-forum at nginx.us> wrote:
> Hi,
>
> I recently installed Debian 6.0.1a Squeeze with Nginx and phpmyadmin
> When I go to 'http://localhost' phpmyadmin pops up.
> What I want is that phpmyadmin is located at
> 'http://localhost/phpmyadmin/'
>
> I already tried to add this in nginx.conf. It didn't work
>
> location /phpmyadmin {
> alias /usr/share/phpmyadmin/;
> }
>
> This is what my nginx.conf looks like:
>
> _____________________________________
> user www-data;
> worker_processes  1;
>
> error_log  /var/log/nginx/error.log;
> pid        /var/run/nginx.pid;
>
> events {
>     worker_connections  1024;
>     # multi_accept on;
> }
>
> http {
>     include       /etc/nginx/mime.types;
>
>     access_log  /var/log/nginx/access.log;
>
>     sendfile        on;
>     #tcp_nopush     on;
>
>     #keepalive_timeout  0;
>     keepalive_timeout  65;
>     tcp_nodelay        on;
>
>     gzip  on;
>     gzip_disable "MSIE [1-6]\.(?!.*SV1)";
>
>     include /etc/nginx/conf.d/*.conf;
>     include /etc/nginx/sites-enabled/*;
> }
> _____________________________________
>
> Can anybody help me configuring this the right way?
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,200488,200488#msg-200488
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>


-- 
--C

"The dumber people think you are, the more surprised they're going to
be when you kill them." - Sir William Clayton


From gvimrc at gmail.com  Wed May 25 03:16:05 2011
From: gvimrc at gmail.com (gvim)
Date: Wed, 25 May 2011 00:16:05 +0100
Subject: Cache problem?
Message-ID: <4DDC3C35.5080400@gmail.com>

I'm running a Perl/PSGI CGI::Application app on nginx via proxy_pass to Starman. The app logs form fields to a database and retrieves a resultset which is served up via Template Toolkit. Running on Apache as a plain vanilla CGI without the PSGI layer I can submit the form, press the Back button when the resultset is returned and re-hit the submit button, all within a second or 2 without any errors. However, with nginx + Starman the same app produces a random number of successful responses then returns an error page from the app as if there were 1 form field too many. It looks like some kind of cached response because even if I remove the module which runs this part of the app I still see the same app-generated error page. Since both nginx and Starman are returning a HTTP body there are no errors in the logs of either daemon to work with.

gvim


From zzz at zzz.org.ua  Wed May 25 03:39:14 2011
From: zzz at zzz.org.ua (Alexandr Gomoliako)
Date: Wed, 25 May 2011 02:39:14 +0300
Subject: Cache problem?
In-Reply-To: <4DDC3C35.5080400@gmail.com>
References: <4DDC3C35.5080400@gmail.com>
Message-ID: <BANLkTikWhCth5wcA1bhrb6WnsQTC8MgHQA@mail.gmail.com>

On 5/25/11, gvim <gvimrc at gmail.com> wrote:
> I'm running a Perl/PSGI CGI::Application app on nginx via proxy_pass to
> Starman. The app logs form fields to a database and retrieves a resultset
> which is served up via Template Toolkit. Running on Apache as a plain
> vanilla CGI without the PSGI layer I can submit the form, press the Back
> button when the resultset is returned and re-hit the submit button, all
> within a second or 2 without any errors. However, with nginx + Starman the
> same app produces a random number of successful responses then returns an
> error page from the app as if there were 1 form field too many. It looks
> like some kind of cached response because even if I remove the module which
> runs this part of the app I still see the same app-generated error page.

It's probably not related to caching whatsoever unless there is a
proxy_cache somewhere in the server block.

> Since both nginx and Starman are returning a HTTP body there are no errors
> in the logs of either daemon to work with.

How about response status?

Anyway, it might be a long shot, but have you tried switching Starman
to a different PSGI server, like Twiggy or Feersum? After all PSGI is
a relatively new thing for perl and not that widely used.


From gvimrc at gmail.com  Wed May 25 04:02:15 2011
From: gvimrc at gmail.com (gvim)
Date: Wed, 25 May 2011 01:02:15 +0100
Subject: Cache problem?
In-Reply-To: <BANLkTikWhCth5wcA1bhrb6WnsQTC8MgHQA@mail.gmail.com>
References: <4DDC3C35.5080400@gmail.com>
	<BANLkTikWhCth5wcA1bhrb6WnsQTC8MgHQA@mail.gmail.com>
Message-ID: <4DDC4707.5030708@gmail.com>

On 25/05/2011 00:39, Alexandr Gomoliako wrote:
>
> How about response status?
>

All 200. The app is returning HTTP content so I don't expect to see any errors on the logs of nginx or Starman.

> Anyway, it might be a long shot, but have you tried switching Starman
> to a different PSGI server, like Twiggy or Feersum? After all PSGI is
> a relatively new thing for perl and not that widely used.

Of the ones you've listed Starman is the more established. My ngxing vhosts file has this kind of setup:

server {
     root /var/www/html/myapp/pub;
     server_name myappserver;
     location ~ \.psgi {
       gzip off;
       proxy_set_header Host $http_host;
       proxy_set_header X-Forwarded-Host $http_host;
       proxy_set_header X-Real-IP $remote_addr;
       proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
       proxy_pass       http://127.0.0.1:9000;
     }
}

... with Starman running on 127.0.0.1:9000

PSGI/Plack is THE big thing in Perl right now so it's actually getting a lot of use, especially within the CGI::Application, Dancer and Catalyst communities.

gvim





From zzz at zzz.org.ua  Wed May 25 04:52:46 2011
From: zzz at zzz.org.ua (Alexandr Gomoliako)
Date: Wed, 25 May 2011 03:52:46 +0300
Subject: Cache problem?
In-Reply-To: <4DDC4707.5030708@gmail.com>
References: <4DDC3C35.5080400@gmail.com>
	<BANLkTikWhCth5wcA1bhrb6WnsQTC8MgHQA@mail.gmail.com>
	<4DDC4707.5030708@gmail.com>
Message-ID: <BANLkTin-3buRG1WAa_qTLsB9jWY3pSe=aA@mail.gmail.com>

On 5/25/11, gvim <gvimrc at gmail.com> wrote:
> Of the ones you've listed Starman is the more established.
> My ngxing vhosts file has this kind of setup:

There is nothing wrong with your nginx setup. So there is a good
chance that either PSGI server or your application does this.


From gvimrc at gmail.com  Wed May 25 05:02:48 2011
From: gvimrc at gmail.com (gvim)
Date: Wed, 25 May 2011 02:02:48 +0100
Subject: Cache problem?
In-Reply-To: <BANLkTin-3buRG1WAa_qTLsB9jWY3pSe=aA@mail.gmail.com>
References: <4DDC3C35.5080400@gmail.com>	<BANLkTikWhCth5wcA1bhrb6WnsQTC8MgHQA@mail.gmail.com>	<4DDC4707.5030708@gmail.com>
	<BANLkTin-3buRG1WAa_qTLsB9jWY3pSe=aA@mail.gmail.com>
Message-ID: <4DDC5538.1050503@gmail.com>

On 25/05/2011 01:52, Alexandr Gomoliako wrote:

>
> There is nothing wrong with your nginx setup. So there is a good
> chance that either PSGI server or your application does this.
>

I just switched it to nginx + FastCGI/cgi-wrap.pl and it ran consistently without errors but was dog slow compared with plain vanilla Apache/CGI. Don't get it.

gvim


From sbc19861004 at gmail.com  Wed May 25 07:38:17 2011
From: sbc19861004 at gmail.com (eagle sbc)
Date: Wed, 25 May 2011 11:38:17 +0800
Subject: SSI include infinite loop
In-Reply-To: <20110524125901.GW42265@mdounin.ru>
References: <BANLkTi=u-pHdYdFA5P2=nAioCuyCija5Cw@mail.gmail.com>
	<20110524125901.GW42265@mdounin.ru>
Message-ID: <BANLkTikJJ4ugjp8YMGuNzmjVXsr-W2bq_w@mail.gmail.com>

Hello,

Thanks for your reply! And I think the limit on the number is proper for
most of the conditions.

However in our case, the index page may have hundreds of SSI 'include'
(combined from lots of our sub sites), and the number limitation may not be
so proper.

So here I made a patch: For every 'include', a 'parents uri check' is made.
The patch is like this:

diff -crB nginx-0.7.65.org/src/http/modules/ngx_http_ssi_filter_module.cnginx-0.7.65.patch/src/http/modules/ngx_http_ssi_filter_module.c
*** nginx-0.7.65.org/src/http/modules/ngx_http_ssi_filter_module.c
2011-05-24 18:46:53.676370002 +0800
--- nginx-0.7.65.patch/src/http/modules/ngx_http_ssi_filter_module.c
2011-05-25 11:19:20.372370002 +0800
***************
*** 2024,2029 ****
--- 2024,2044 ----
          flags |= NGX_HTTP_SUBREQUEST_IN_MEMORY|NGX_HTTP_SUBREQUEST_WAITED;
      }

+     ngx_http_request_t *ir = r;
+     for(;ir != NULL;ir = ir->parent) {
+         ngx_log_debug4(NGX_LOG_DEBUG_HTTP, r->connection->log, 0,
+                        " include path: %V, args: %V; parent path: %V,
args: %V",
+                        uri, &args, &ir->uri, &ir->args);
+         if (uri->len == ir->uri.len
+             && args.len == ir->args.len
+             && ngx_strncmp(uri->data, ir->uri.data, uri->len) == 0
+             && ngx_strncmp(args.data, ir->args.data, args.len) == 0) {
+             ngx_log_error(NGX_LOG_ERR, r->connection->log, 0,
+                           "A loop include \"%V\" is found", uri);
+             return NGX_HTTP_SSI_ERROR;
+         }
+     }
+
      if (ngx_http_subrequest(r, uri, &args, &sr, psr, flags) != NGX_OK) {
          return NGX_HTTP_SSI_ERROR;
      }

Hope for any reply of my patch !

And thanks to my fellow ??? for reminding me on the 'args' check!


On Tue, May 24, 2011 at 8:59 PM, Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
>
> On Tue, May 24, 2011 at 10:49:45AM +0800, eagle sbc wrote:
>
> > Hello,
> >
> > I've encountered a problem about "SSI include infinite loop".
> >
> > We have a system that our editors can compose a html page from some
> > templates using SSI. Sometimes unwarily combination will cause infinite
> > loop, such as 'A include A', 'A include B, B include A', or even 'A
> include
> > B, B include C, C include A'. And this page will cause nginx eating up
> all
> > memories, and certainly won't serve any more unless we restart nginx.
> >
> > I'm wondering is there any solution for this problem? Or other ways to
> avoid
> > this?
> >
> > I've googled apache and lighttpd, and there also no official solution for
> > this, unless lightptd has a patch which will limit the whole 'include'
> > count, and I think it's just a tricky way to fix the infinite loop.
> >
> > Thanks a lot for any help!
>
> Number of in-flight ssi includes used to be limited in nginx,
> though this limit was broken somewhere near 0.7.25.
>
> This patch brings it back:
>
> http://nginx.org/pipermail/nginx-devel/2011-February/000746.html
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>



-- 
???????????????????
??????????????????????????
???????
??????????????
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110525/d001b274/attachment-0001.html>

From jeffrpang at gmail.com  Wed May 25 09:10:54 2011
From: jeffrpang at gmail.com (Jeff Pang)
Date: Wed, 25 May 2011 13:10:54 +0800
Subject: clean nginx log
Message-ID: <BANLkTi=CVzoWPHF6mEDdof=f=8dtpacGBQ@mail.gmail.com>

Hello,

I'm not sure, is it just right to truncate the nginx's access log by:

echo -n > /path/to/log/access.log

Thanks.

-- 
Jeff Pang
www.DNSbed.com


From igor at sysoev.ru  Wed May 25 09:38:17 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 09:38:17 +0400
Subject: clean nginx log
In-Reply-To: <BANLkTi=CVzoWPHF6mEDdof=f=8dtpacGBQ@mail.gmail.com>
References: <BANLkTi=CVzoWPHF6mEDdof=f=8dtpacGBQ@mail.gmail.com>
Message-ID: <20110525053817.GD59490@sysoev.ru>

On Wed, May 25, 2011 at 01:10:54PM +0800, Jeff Pang wrote:
> Hello,
> 
> I'm not sure, is it just right to truncate the nginx's access log by:
> 
> echo -n > /path/to/log/access.log

nginx uses O_APPEND flag for access log, so this trick may work.
There is however, a more safer way:

rm access.log
kill -USR1 `cat /path/to/nginx.pid`


-- 
Igor Sysoev


From igor at sysoev.ru  Wed May 25 10:13:48 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 10:13:48 +0400
Subject: Provide site-specific SSL cert on behalf of clients
In-Reply-To: <bb893cda9c86fc398927edad02f87284.NginxMailingListEnglish@forum.nginx.org>
References: <bb893cda9c86fc398927edad02f87284.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525061348.GF59490@sysoev.ru>

On Tue, May 24, 2011 at 05:04:11PM -0400, urschrei wrote:
> I'm wondering if I can configure nginx to provide a PEM SSL cert for a
> specific site on behalf of clients on my subnet, and how I might go
> about configuring, testing this. I have a cert in PEM format, and nginx
> 0.8.54 with SSL support, running on Ubuntu 11.04.

Do you mean a cetificate that a client passes to a server or what ?


-- 
Igor Sysoev


From jeffrpang at gmail.com  Wed May 25 10:24:40 2011
From: jeffrpang at gmail.com (Jeff Pang)
Date: Wed, 25 May 2011 14:24:40 +0800
Subject: clean nginx log
In-Reply-To: <20110525053817.GD59490@sysoev.ru>
References: <BANLkTi=CVzoWPHF6mEDdof=f=8dtpacGBQ@mail.gmail.com>
	<20110525053817.GD59490@sysoev.ru>
Message-ID: <BANLkTik9t9j=bgx-zZvM9kXRyDfCmsYHeA@mail.gmail.com>

2011/5/25 Igor Sysoev <igor at sysoev.ru>:

>
> nginx uses O_APPEND flag for access log, so this trick may work.
> There is however, a more safer way:
>
> rm access.log
> kill -USR1 `cat /path/to/nginx.pid`
>
>

That's right. Thanks Igor.

-- 
Jeff Pang
www.DNSbed.com


From nginx-forum at nginx.us  Wed May 25 13:15:18 2011
From: nginx-forum at nginx.us (urschrei)
Date: Wed, 25 May 2011 05:15:18 -0400
Subject: Provide site-specific SSL cert on behalf of clients
In-Reply-To: <20110525061348.GF59490@sysoev.ru>
References: <20110525061348.GF59490@sysoev.ru>
Message-ID: <98641453b4bf34b5f9f3b17d98af2f22.NginxMailingListEnglish@forum.nginx.org>

Igor,
yes, it's a client SSL cert.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200521,200646#msg-200646



From igor at sysoev.ru  Wed May 25 13:19:57 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 13:19:57 +0400
Subject: Provide site-specific SSL cert on behalf of clients
In-Reply-To: <98641453b4bf34b5f9f3b17d98af2f22.NginxMailingListEnglish@forum.nginx.org>
References: <20110525061348.GF59490@sysoev.ru>
	<98641453b4bf34b5f9f3b17d98af2f22.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525091957.GB83054@sysoev.ru>

On Wed, May 25, 2011 at 05:15:18AM -0400, urschrei wrote:
> Igor,
> yes, it's a client SSL cert.

You should have a CA certificate which is used to sign client certificates.
This ceritificate should be declared on server level:

server {
    listen  443;
    ssl     on;
    ...

    ssl_client_certificate  ca-cert.pem;
    ssl_verify_client       on;


-- 
Igor Sysoev


From nginx-forum at nginx.us  Wed May 25 13:41:14 2011
From: nginx-forum at nginx.us (Trazzt)
Date: Wed, 25 May 2011 05:41:14 -0400
Subject: Configuring phpmyadmin
In-Reply-To: <BANLkTi=_Cr46Dktt6kuuT=it6fMKBAFkYg@mail.gmail.com>
References: <BANLkTi=_Cr46Dktt6kuuT=it6fMKBAFkYg@mail.gmail.com>
Message-ID: <c1efdb2f33c60a4bbd8b91ba06657f41.NginxMailingListEnglish@forum.nginx.org>

Okay thx for the tip I'm new to this ;)
Here it is:

phpmyadmin:
_____________________________________
server{
    listen 80;
    server_name phpmyadmin;
    access_log /var/log/phpmyadmin.access_log;
    error_log /var/log/phpmyadmin.error_log;
 
    location / {
      root /usr/share/phpmyadmin;
      index index.php;
    }
 
    location ~ \.php$ {
        fastcgi_pass    127.0.0.1:9000;
        fastcgi_index   index.php;
        fastcgi_param   SCRIPT_FILENAME
/usr/share/phpmyadmin$fastcgi_script_name;
        include         fastcgi_params;
    }
}
_____________________________________

This is my fastcgi_script, located at "/etc/init.d/php-fastcgi"
http://pastie.org/1970419


I also have one other site enabled, I don't know if thats relevant but
you can find it here.
I want this site to be located at 'localhost'
http://pastie.org/1970411

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200488,200656#msg-200656



From igor at sysoev.ru  Wed May 25 13:45:48 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 13:45:48 +0400
Subject: Configuring phpmyadmin
In-Reply-To: <c1efdb2f33c60a4bbd8b91ba06657f41.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTi=_Cr46Dktt6kuuT=it6fMKBAFkYg@mail.gmail.com>
	<c1efdb2f33c60a4bbd8b91ba06657f41.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525094548.GD83054@sysoev.ru>

On Wed, May 25, 2011 at 05:41:14AM -0400, Trazzt wrote:
> Okay thx for the tip I'm new to this ;)
> Here it is:
> 
> phpmyadmin:
> _____________________________________
> server{
>     listen 80;
>     server_name phpmyadmin;
>     access_log /var/log/phpmyadmin.access_log;
>     error_log /var/log/phpmyadmin.error_log;
>  
>     location / {
>       root /usr/share/phpmyadmin;
>       index index.php;
>     }
>  
>     location ~ \.php$ {
>         fastcgi_pass    127.0.0.1:9000;
>         fastcgi_index   index.php;
>         fastcgi_param   SCRIPT_FILENAME
> /usr/share/phpmyadmin$fastcgi_script_name;
>         include         fastcgi_params;
>     }
> }
> _____________________________________
> 
> This is my fastcgi_script, located at "/etc/init.d/php-fastcgi"
> http://pastie.org/1970419
> 
> 
> I also have one other site enabled, I don't know if thats relevant but
> you can find it here.
> I want this site to be located at 'localhost'
> http://pastie.org/1970411

You probably need

    location / {
        return 404;
    }

    location /phpmyadmin {
      root /usr/share;
      index index.php;
    }
 
    location ~ ^/phpmyadmin/\.php$ {
        fastcgi_pass    127.0.0.1:9000;
        fastcgi_param   SCRIPT_FILENAME /usr/share$fastcgi_script_name;
        include         fastcgi_params;
    }


-- 
Igor Sysoev


From nginx-forum at nginx.us  Wed May 25 13:48:26 2011
From: nginx-forum at nginx.us (speedfirst)
Date: Wed, 25 May 2011 05:48:26 -0400
Subject: How to add an arg to the requests to the upstreams
Message-ID: <d4ef614930eba55d01ca78de4aa31218.NginxMailingListEnglish@forum.nginx.org>

Hey, I wanna add a custom arg to the upstreams by proxy_pass like this:

upstream servers {
    server1:8011;
    server2:8012;
 }

...
server {
  listen on 8443;
  ssl on;
  server_name server0;
  location / {
     proxy_pass https://servers; 
  }
}


My expectation: type "https://server0:8443" and the upstream will get
"https://server1:8011?abc=1"

So how can I make this? Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200661,200661#msg-200661



From igor at sysoev.ru  Wed May 25 13:54:25 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 13:54:25 +0400
Subject: How to add an arg to the requests to the upstreams
In-Reply-To: <d4ef614930eba55d01ca78de4aa31218.NginxMailingListEnglish@forum.nginx.org>
References: <d4ef614930eba55d01ca78de4aa31218.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525095425.GE83054@sysoev.ru>

On Wed, May 25, 2011 at 05:48:26AM -0400, speedfirst wrote:
> Hey, I wanna add a custom arg to the upstreams by proxy_pass like this:
> 
> upstream servers {
>     server1:8011;
>     server2:8012;
>  }
> 
> ...
> server {
>   listen on 8443;
>   ssl on;
>   server_name server0;
>   location / {
>      proxy_pass https://servers; 
>   }
> }
> 
> 
> My expectation: type "https://server0:8443" and the upstream will get
> "https://server1:8011?abc=1"
> 
> So how can I make this? Thanks.

      proxy_pass  https://servers$uri?abc=1;


-- 
Igor Sysoev


From nginx-forum at nginx.us  Wed May 25 14:12:03 2011
From: nginx-forum at nginx.us (speedfirst)
Date: Wed, 25 May 2011 06:12:03 -0400
Subject: How to add an arg to the requests to the upstreams
In-Reply-To: <d4ef614930eba55d01ca78de4aa31218.NginxMailingListEnglish@forum.nginx.org>
References: <d4ef614930eba55d01ca78de4aa31218.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <3746305b0f4b3659bdda38df5bdcd362.NginxMailingListEnglish@forum.nginx.org>

Is direct pass OK, like this?

proxy_pass https://servers?abc=1

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200661,200666#msg-200666



From nginx-forum at nginx.us  Wed May 25 14:13:49 2011
From: nginx-forum at nginx.us (Trazzt)
Date: Wed, 25 May 2011 06:13:49 -0400
Subject: Configuring phpmyadmin
In-Reply-To: <20110525094548.GD83054@sysoev.ru>
References: <20110525094548.GD83054@sysoev.ru>
Message-ID: <1a34a17cc38e0ae28d7d2f29cbaad85f.NginxMailingListEnglish@forum.nginx.org>

Thank you, that did the trick!
Only I have a new problem now. PHP looks suddenly disabled
I get a text file, when I visit: '127.0.0.1/phpmyadmin'
Like this:
 http://postimage.org/image/2j6o8olyc/

I already tried to add the line:
        fastcgi_index   index.php;
and restarting nginx. That didn't work
Rebooting also didn't help

Do you know how I can enable PHP again?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200488,200667#msg-200667



From igor at sysoev.ru  Wed May 25 14:25:21 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 14:25:21 +0400
Subject: Configuring phpmyadmin
In-Reply-To: <1a34a17cc38e0ae28d7d2f29cbaad85f.NginxMailingListEnglish@forum.nginx.org>
References: <20110525094548.GD83054@sysoev.ru>
	<1a34a17cc38e0ae28d7d2f29cbaad85f.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525102521.GF83054@sysoev.ru>

On Wed, May 25, 2011 at 06:13:49AM -0400, Trazzt wrote:
> Thank you, that did the trick!
> Only I have a new problem now. PHP looks suddenly disabled
> I get a text file, when I visit: '127.0.0.1/phpmyadmin'
> Like this:
>  http://postimage.org/image/2j6o8olyc/
> 
> I already tried to add the line:
>         fastcgi_index   index.php;
> and restarting nginx. That didn't work

You do not need this directive.

> Rebooting also didn't help
> 
> Do you know how I can enable PHP again?

-    location ~ ^/phpmyadmin/\.php$ {
+    location ~ ^/phpmyadmin/.+\.php$ {

This is yet one of the reasons why I try to avoid as much as possible
regex locations - too high mistake probability if you do not write or read
a hundrend regexes a day.


-- 
Igor Sysoev


From igor at sysoev.ru  Wed May 25 14:25:37 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 14:25:37 +0400
Subject: How to add an arg to the requests to the upstreams
In-Reply-To: <3746305b0f4b3659bdda38df5bdcd362.NginxMailingListEnglish@forum.nginx.org>
References: <d4ef614930eba55d01ca78de4aa31218.NginxMailingListEnglish@forum.nginx.org>
	<3746305b0f4b3659bdda38df5bdcd362.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525102537.GG83054@sysoev.ru>

On Wed, May 25, 2011 at 06:12:03AM -0400, speedfirst wrote:
> Is direct pass OK, like this?
> 
> proxy_pass https://servers?abc=1

No.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Wed May 25 14:27:15 2011
From: nginx-forum at nginx.us (urschrei)
Date: Wed, 25 May 2011 06:27:15 -0400
Subject: Provide site-specific SSL cert on behalf of clients
In-Reply-To: <20110525091957.GB83054@sysoev.ru>
References: <20110525091957.GB83054@sysoev.ru>
Message-ID: <1d3873fa52a49afa95a44b671017be50.NginxMailingListEnglish@forum.nginx.org>

Igor,
just to make sure I'm not misunderstanding you:

Usually, what happens is this:

I install an SSL cert (let's call it certA) in a client browser, so I
can access https site A, which requires it.

But if I have a lot of clients, I'd ideally like to have nginx proxy
this cert, on behalf of my clients, so I don't have to install it for
each of them. Are you saying that in order for nginx to proxy the cert,
I'll first have to generate a CA cert on the server, and then sign the
client cert (certA) with it? Won't this result in a self-signed
certificate warning every time a client tries to access site A?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200521,200672#msg-200672



From igor at sysoev.ru  Wed May 25 14:34:03 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 14:34:03 +0400
Subject: Provide site-specific SSL cert on behalf of clients
In-Reply-To: <1d3873fa52a49afa95a44b671017be50.NginxMailingListEnglish@forum.nginx.org>
References: <20110525091957.GB83054@sysoev.ru>
	<1d3873fa52a49afa95a44b671017be50.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525103403.GH83054@sysoev.ru>

On Wed, May 25, 2011 at 06:27:15AM -0400, urschrei wrote:
> Igor,
> just to make sure I'm not misunderstanding you:
> 
> Usually, what happens is this:
> 
> I install an SSL cert (let's call it certA) in a client browser, so I
> can access https site A, which requires it.
> 
> But if I have a lot of clients, I'd ideally like to have nginx proxy
> this cert, on behalf of my clients, so I don't have to install it for
> each of them. Are you saying that in order for nginx to proxy the cert,
> I'll first have to generate a CA cert on the server, and then sign the
> client cert (certA) with it? Won't this result in a self-signed
> certificate warning every time a client tries to access site A?

nginx as a client does not currently support a client certificate
when it proxies a request to HTTPS backend (B). However, I do not see
any security advantage when many clients look like one for backend B.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Wed May 25 14:46:17 2011
From: nginx-forum at nginx.us (urschrei)
Date: Wed, 25 May 2011 06:46:17 -0400
Subject: Provide site-specific SSL cert on behalf of clients
In-Reply-To: <20110525103403.GH83054@sysoev.ru>
References: <20110525103403.GH83054@sysoev.ru>
Message-ID: <ca5cd4c4f600806ad6540b6de7687610.NginxMailingListEnglish@forum.nginx.org>

> nginx as a client does not currently support a
> client certificate
> when it proxies a request to HTTPS backend (B).
> However, I do not see
> any security advantage when many clients look like
> one for backend B.

Thanks for clearing up the confusion. In this case, there would be no
security advantage even if the cert were individually installed, since
the same cert is issued to each client in the organisation, so whether
the cert is proxied or not really makes no difference.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200521,200678#msg-200678



From igor at sysoev.ru  Wed May 25 15:37:08 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 15:37:08 +0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <eaddf429fe128da4d46fcde97e1ddf24.NginxMailingListEnglish@forum.nginx.org>
References: <c0c6df7135604bcee44a098e5b896a30@ruby-forum.com>
	<e00b353ec548b2c91b169c5b8ccaaa9c.NginxMailingListEnglish@forum.nginx.org>
	<6856e4d76e54e6ffbc66eaa7c05442ce.NginxMailingListEnglish@forum.nginx.org>
	<4afe9fa1984e2ca74c7636d2a379d417.NginxMailingListEnglish@forum.nginx.org>
	<eaddf429fe128da4d46fcde97e1ddf24.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525113708.GJ83054@sysoev.ru>

On Mon, May 23, 2011 at 01:18:30PM -0400, the_traveller wrote:
> Hi mr Igor 
> I'm sory coz i toke alittle bit too long
> 
> so if i have to make any ngnix rule i have to make it like that by ex:
> 
> location /Joke/ {
> fastcgi_pass ...
> fastcgi_param SCRIPT_FILENAME /jokes.php;
> fastcgi_param QUERY_STRING $args;
> ...
> }
> 
> and the Original Apache rule is
> RewriteRule ^Joke/.*     jokes.php       [QSA]
> 
> sposing that the script is in the public dir  of the ftp 
> is my try right??

SCRIPT_FILENAME should point to a file, for example,
   /usr/www/scripts/jokes.php.

> and also is the three dots in the second line
> fastcgi_pass ...
> and before the closing tag
> ...
> }
>  is this is right or i have to remove them ??
> 
> thanks alot for ur help
> and if i have removed them

dots means that you should add your specific paramters, that is,

     fastcgi_pass   backend:9000;

etc.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Wed May 25 15:43:54 2011
From: nginx-forum at nginx.us (chris.percol)
Date: Wed, 25 May 2011 07:43:54 -0400
Subject: redirect subdomain to internal location
Message-ID: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>

How would I test for the following subdomains?

www
www4
any other subdomain 

Then redirect each to a corresponding internal location?

I am using the following code to get the 'any other subdomain' but can't
get the redirect to a  location such as @any_subd working?

if ($host ~* "(.*)\.tm2app\.com") {
  set $subd $1;
}

Thanks for your help.

Chris

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200694,200694#msg-200694



From nginx-forum at nginx.us  Wed May 25 15:58:21 2011
From: nginx-forum at nginx.us (Trazzt)
Date: Wed, 25 May 2011 07:58:21 -0400
Subject: Configuring phpmyadmin
In-Reply-To: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
References: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <2939c25028e7d77fdd6ffa5fcf62a654.NginxMailingListEnglish@forum.nginx.org>

I really appreciate your help, but PHP is still not working
This is how my phpmyadmin configuration file looks like after all your
edits:
________________________
server{
    listen 80;
    server_name phpmyadmin;
    access_log /var/log/phpmyadmin.access_log;
    error_log /var/log/phpmyadmin.error_log;
 
    location / {
    return 404;
      }

    location /phpmyadmin {
    root /usr/share;
    index index.php;
      }

    location ~ ^/phpmyadmin/.+\.php$ {
    fastcgi_pass 127.0.0.1:9000;
    fastcgi_param SCRIPT_FILENAME /usr/share$fastcgi_script_name;
    include fastcgi_params;
      }
  }
______________________

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200488,200702#msg-200702



From appa at perusio.net  Wed May 25 15:57:54 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Wed, 25 May 2011 12:57:54 +0100
Subject: redirect subdomain to internal location
In-Reply-To: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>
References: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87d3j7kn5p.wl%appa@perusio.net>

On 25 Mai 2011 12h43 WEST, nginx-forum at nginx.us wrote:

> How would I test for the following subdomains?
>
> www
> www4
> any other subdomain 
>
> Then redirect each to a corresponding internal location?
>
> I am using the following code to get the 'any other subdomain' but
> can't get the redirect to a location such as @any_subd working?
>
> if ($host ~* "(.*)\.tm2app\.com") {
> set $subd $1;
> }

Do you mean that *all* subdomains should use the same location?

If so use a wildcard in the server_name directive:

server {
  server_name *.tmp2app.com;

  (...)
}

If the subdomains always start by www and they're to be handled
differently than other do this for www1, www2...www99

server {
  server_name "~^www\d{1,2}\.tmp2app\.com$";
  (...)
}

--- appa



> Thanks for your help.
>
> Chris
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,200694,200694#msg-200694
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx


From dick at fouter.net  Wed May 25 16:04:44 2011
From: dick at fouter.net (Dick Middleton)
Date: Wed, 25 May 2011 13:04:44 +0100
Subject: Configuring phpmyadmin
In-Reply-To: <2939c25028e7d77fdd6ffa5fcf62a654.NginxMailingListEnglish@forum.nginx.org>
References: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
	<2939c25028e7d77fdd6ffa5fcf62a654.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4DDCF05C.4010807@fouter.net>

On 05/25/11 12:58, Trazzt wrote:
> I really appreciate your help, but PHP is still not working
> This is how my phpmyadmin configuration file looks like after all your
> edits:

>      location ~ ^/phpmyadmin/.+\.php$ {
>      fastcgi_pass 127.0.0.1:9000;
>      fastcgi_param SCRIPT_FILENAME /usr/share$fastcgi_script_name;
>      include fastcgi_params;

I found I had to use:

fastcgi_param  SCRIPT_FILENAME  $request_filename;

I don't really understand why.

Dick


From igor at sysoev.ru  Wed May 25 16:48:50 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 16:48:50 +0400
Subject: Configuring phpmyadmin
In-Reply-To: <4DDCF05C.4010807@fouter.net>
References: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
	<2939c25028e7d77fdd6ffa5fcf62a654.NginxMailingListEnglish@forum.nginx.org>
	<4DDCF05C.4010807@fouter.net>
Message-ID: <20110525124850.GL83054@sysoev.ru>

On Wed, May 25, 2011 at 01:04:44PM +0100, Dick Middleton wrote:
> On 05/25/11 12:58, Trazzt wrote:
> > I really appreciate your help, but PHP is still not working
> > This is how my phpmyadmin configuration file looks like after all your
> > edits:
> 
> >      location ~ ^/phpmyadmin/.+\.php$ {
> >      fastcgi_pass 127.0.0.1:9000;
> >      fastcgi_param SCRIPT_FILENAME /usr/share$fastcgi_script_name;
> >      include fastcgi_params;
> 
> I found I had to use:
> 
> fastcgi_param  SCRIPT_FILENAME  $request_filename;
> 
> I don't really understand why.

$request_filename is "$document_root$uri".


-- 
????? ??????
http://sysoev.ru


From igor at sysoev.ru  Wed May 25 16:50:03 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 16:50:03 +0400
Subject: Configuring phpmyadmin
In-Reply-To: <2939c25028e7d77fdd6ffa5fcf62a654.NginxMailingListEnglish@forum.nginx.org>
References: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
	<2939c25028e7d77fdd6ffa5fcf62a654.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525125003.GN83054@sysoev.ru>

On Wed, May 25, 2011 at 07:58:21AM -0400, Trazzt wrote:
> I really appreciate your help, but PHP is still not working

What does it means - "PHP is still not working" ?

> This is how my phpmyadmin configuration file looks like after all your
> edits:
> ________________________
> server{
>     listen 80;
>     server_name phpmyadmin;
>     access_log /var/log/phpmyadmin.access_log;
>     error_log /var/log/phpmyadmin.error_log;
>  
>     location / {
>     return 404;
>       }
> 
>     location /phpmyadmin {
>     root /usr/share;
>     index index.php;
>       }
> 
>     location ~ ^/phpmyadmin/.+\.php$ {
>     fastcgi_pass 127.0.0.1:9000;
>     fastcgi_param SCRIPT_FILENAME /usr/share$fastcgi_script_name;
>     include fastcgi_params;
>       }
>   }
> ______________________


-- 
????? ??????
http://sysoev.ru


From nginx-forum at nginx.us  Wed May 25 17:07:30 2011
From: nginx-forum at nginx.us (Trazzt)
Date: Wed, 25 May 2011 09:07:30 -0400
Subject: Configuring phpmyadmin
In-Reply-To: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
References: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <8ceb852c72ec1949f2d94c7ffe484bd9.NginxMailingListEnglish@forum.nginx.org>

I mean I still get the text screen like in the image. I still get no
layout, only the 'index.php' file in text format.
Like in the screenshot:
http://postimage.org/image/2j6o8olyc/

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200488,200733#msg-200733



From igor at sysoev.ru  Wed May 25 17:15:57 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 17:15:57 +0400
Subject: Configuring phpmyadmin
In-Reply-To: <8ceb852c72ec1949f2d94c7ffe484bd9.NginxMailingListEnglish@forum.nginx.org>
References: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
	<8ceb852c72ec1949f2d94c7ffe484bd9.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525131557.GR83054@sysoev.ru>

On Wed, May 25, 2011 at 09:07:30AM -0400, Trazzt wrote:
> I mean I still get the text screen like in the image. I still get no
> layout, only the 'index.php' file in text format.
> Like in the screenshot:
> http://postimage.org/image/2j6o8olyc/

Try to add localhost and 127.0.0.1 to ensure that you request a right
server:

server{
    listen 80;
    server_name phpmyadmin localhost 127.0.0.1;


-- 
Igor Sysoev


From nginx-forum at nginx.us  Wed May 25 17:16:31 2011
From: nginx-forum at nginx.us (chris.percol)
Date: Wed, 25 May 2011 09:16:31 -0400
Subject: redirect subdomain to internal location
In-Reply-To: <87d3j7kn5p.wl%appa@perusio.net>
References: <87d3j7kn5p.wl%appa@perusio.net>
Message-ID: <af255a1d6b59cbe20d338d3b2336aa8e.NginxMailingListEnglish@forum.nginx.org>

Thanks.

No, I want www and ww4 to use different locations, with all other
subdomains using another single location.

Thanks again,

Chris

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200694,200739#msg-200739



From igor at sysoev.ru  Wed May 25 17:21:00 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 17:21:00 +0400
Subject: redirect subdomain to internal location
In-Reply-To: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>
References: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525132100.GS83054@sysoev.ru>

On Wed, May 25, 2011 at 07:43:54AM -0400, chris.percol wrote:
> How would I test for the following subdomains?
> 
> www
> www4
> any other subdomain 
> 
> Then redirect each to a corresponding internal location?
> 
> I am using the following code to get the 'any other subdomain' but can't
> get the redirect to a  location such as @any_subd working?
> 
> if ($host ~* "(.*)\.tm2app\.com") {
>   set $subd $1;
> }

Could you describe using examples, but not "if/set" ?
nginx virtual server names are described here:
http://nginx.org/en/docs/http/server_names.html


-- 
Igor Sysoev


From appa at perusio.net  Wed May 25 17:22:50 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Wed, 25 May 2011 14:22:50 +0100
Subject: redirect subdomain to internal location
In-Reply-To: <af255a1d6b59cbe20d338d3b2336aa8e.NginxMailingListEnglish@forum.nginx.org>
References: <87d3j7kn5p.wl%appa@perusio.net>
	<af255a1d6b59cbe20d338d3b2336aa8e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87boyqlxsl.wl%appa@perusio.net>

On 25 Mai 2011 14h16 WEST, nginx-forum at nginx.us wrote:

> Thanks.
>
> No, I want www and ww4 to use different locations, with all other
> subdomains using another single location.

server {
  server_name www.tmp2app.com;

  (...) # put one location here
}

server {
  server_name www4.tmp2app.com;
  
  (...) # put the other location here
}

server {
  server_name *.tmp2app.com;
  
  (...) # put the remaining here
}

--- appa



From nginx-forum at nginx.us  Wed May 25 17:34:57 2011
From: nginx-forum at nginx.us (Trazzt)
Date: Wed, 25 May 2011 09:34:57 -0400
Subject: Configuring phpmyadmin
In-Reply-To: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
References: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <2ece37602602a6f14d386e970ca800fe.NginxMailingListEnglish@forum.nginx.org>

I tried what you said, but I still get the text file.

Thanks for all your support, I'm gonna reinstall phpmyadmin completely
using the configuration files you supplied.

I used this tutorial to install nginx + phpmyadmin, I'll search for
another tut.
http://joneslee85.wordpress.com/2010/02/28/howto-nginx-php5-mysql-phpmyadmin-ubuntu-shortest-setup/

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200488,200751#msg-200751



From igor at sysoev.ru  Wed May 25 17:36:29 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 17:36:29 +0400
Subject: Configuring phpmyadmin
In-Reply-To: <2ece37602602a6f14d386e970ca800fe.NginxMailingListEnglish@forum.nginx.org>
References: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
	<2ece37602602a6f14d386e970ca800fe.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525133629.GT83054@sysoev.ru>

On Wed, May 25, 2011 at 09:34:57AM -0400, Trazzt wrote:
> I tried what you said, but I still get the text file.
> 
> Thanks for all your support, I'm gonna reinstall phpmyadmin completely
> using the configuration files you supplied.
> 
> I used this tutorial to install nginx + phpmyadmin, I'll search for
> another tut.
> http://joneslee85.wordpress.com/2010/02/28/howto-nginx-php5-mysql-phpmyadmin-ubuntu-shortest-setup/

Could you create debug log of the request ?
http://nginx.org/en/docs/debugging_log.html


-- 
Igor Sysoev


From dick at fouter.net  Wed May 25 17:56:07 2011
From: dick at fouter.net (Dick Middleton)
Date: Wed, 25 May 2011 14:56:07 +0100
Subject: Configuring phpmyadmin
In-Reply-To: <2939c25028e7d77fdd6ffa5fcf62a654.NginxMailingListEnglish@forum.nginx.org>
References: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
	<2939c25028e7d77fdd6ffa5fcf62a654.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4DDD0A77.9000208@fouter.net>

On 05/25/11 12:58, Trazzt wrote:
> I really appreciate your help, but PHP is still not working
> This is how my phpmyadmin configuration file looks like after all your
> edits:
> ________________________
> server{
>      listen 80;
>      server_name phpmyadmin;
>      access_log /var/log/phpmyadmin.access_log;
>      error_log /var/log/phpmyadmin.error_log;
>
>      location / {
>      return 404;
>        }
>
>      location /phpmyadmin {
>      root /usr/share;
>      index index.php;
>        }
>
>      location ~ ^/phpmyadmin/.+\.php$ {
>      fastcgi_pass 127.0.0.1:9000;
>      fastcgi_param SCRIPT_FILENAME /usr/share$fastcgi_script_name;
>      include fastcgi_params;
>        }
>    }

I'm not sure which location /phpmyadmin/ is served from.  If it's the 
first, as I suspect, then there's no way index.php is going to be processed.

I would use:

     location /phpmyadmin {
	index index.php;
         ...

        location  ~ \.php$ {
        ...
        }
     }

No?

Dick




From igor at sysoev.ru  Wed May 25 18:00:48 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 18:00:48 +0400
Subject: Configuring phpmyadmin
In-Reply-To: <4DDD0A77.9000208@fouter.net>
References: <803a4e8efdcc2152348d257454089b78.NginxMailingListEnglish@forum.nginx.org>
	<2939c25028e7d77fdd6ffa5fcf62a654.NginxMailingListEnglish@forum.nginx.org>
	<4DDD0A77.9000208@fouter.net>
Message-ID: <20110525140048.GU83054@sysoev.ru>

On Wed, May 25, 2011 at 02:56:07PM +0100, Dick Middleton wrote:
> On 05/25/11 12:58, Trazzt wrote:
> > I really appreciate your help, but PHP is still not working
> > This is how my phpmyadmin configuration file looks like after all your
> > edits:
> > ________________________
> > server{
> >      listen 80;
> >      server_name phpmyadmin;
> >      access_log /var/log/phpmyadmin.access_log;
> >      error_log /var/log/phpmyadmin.error_log;
> >
> >      location / {
> >      return 404;
> >        }
> >
> >      location /phpmyadmin {
> >      root /usr/share;
> >      index index.php;
> >        }
> >
> >      location ~ ^/phpmyadmin/.+\.php$ {
> >      fastcgi_pass 127.0.0.1:9000;
> >      fastcgi_param SCRIPT_FILENAME /usr/share$fastcgi_script_name;
> >      include fastcgi_params;
> >        }
> >    }
> 
> I'm not sure which location /phpmyadmin/ is served from.  If it's the 
> first, as I suspect, then there's no way index.php is going to be processed.
> 
> I would use:
> 
>      location /phpmyadmin {
> 	index index.php;
>          ...
> 
>         location  ~ \.php$ {
>         ...
>         }
>      }
> 
> No?

Yes, inclusive "location  ~ \.php$" is much better than configuration
I suggested.

However, my configuration should also work, since /phpmyadmin/ is
internally redirected to /phpmyadmin/index.php, see for details:
http://nginx.org/en/docs/http/request_processing.html#simple_php_site_configuration


-- 
Igor Sysoev


From nginx-forum at nginx.us  Wed May 25 18:26:42 2011
From: nginx-forum at nginx.us (chris.percol)
Date: Wed, 25 May 2011 10:26:42 -0400
Subject: redirect subdomain to internal location
In-Reply-To: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>
References: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <30128c32f2b4ce211f2854f9ef451967.NginxMailingListEnglish@forum.nginx.org>

Thanks guys,

I get that now, in my newness to nginx I was overcomplicating things. My
server_name is now doing what I had hoped to achieve.

If you can bear with me for one more question, how do I get the 'client'
out of domain.com/client/more/stuff.htm? I want to use the 'client' part
of the uri to search a redis key/value pair.

I had success with this code for query strings but couldn't work out how
to get the first part of the uri to be my key?

       location / {
       eval_escalate on;
       eval $res {
               set_unescape_uri $key $arg_key;
               redis2_query get $key;
               redis2_pass 127.0.0.1:6379;
       }
       set_by_lua $path "return
require('redis.parser').parse_reply(ngx.arg[1])" $res;
       proxy_pass http://app_cluster_1/$path;
       proxy_redirect off;
       }


Thanks again for all the help!

Chris

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200694,200774#msg-200774



From nginx-forum at nginx.us  Wed May 25 18:38:56 2011
From: nginx-forum at nginx.us (Trazzt)
Date: Wed, 25 May 2011 10:38:56 -0400
Subject: Configuring phpmyadmin
In-Reply-To: <20110525140048.GU83054@sysoev.ru>
References: <20110525140048.GU83054@sysoev.ru>
Message-ID: <5fbdb16d9c5bea48f084b81fc00a688d.NginxMailingListEnglish@forum.nginx.org>

Igor Sysoev Wrote:
-------------------------------------------------------
> On Wed, May 25, 2011 at 02:56:07PM +0100, Dick
> Middleton wrote:
> > On 05/25/11 12:58, Trazzt wrote:
> > > I really appreciate your help, but PHP is
> still not working
> > > This is how my phpmyadmin configuration file
> looks like after all your
> > > edits:
> > > ________________________
> > > server{
> > >      listen 80;
> > >      server_name phpmyadmin;
> > >      access_log
> /var/log/phpmyadmin.access_log;
> > >      error_log /var/log/phpmyadmin.error_log;
> > >
> > >      location / {
> > >      return 404;
> > >        }
> > >
> > >      location /phpmyadmin {
> > >      root /usr/share;
> > >      index index.php;
> > >        }
> > >
> > >      location ~ ^/phpmyadmin/.+\.php$ {
> > >      fastcgi_pass 127.0.0.1:9000;
> > >      fastcgi_param SCRIPT_FILENAME
> /usr/share$fastcgi_script_name;
> > >      include fastcgi_params;
> > >        }
> > >    }
> > 
> > I'm not sure which location /phpmyadmin/ is
> served from.  If it's the 
> > first, as I suspect, then there's no way
> index.php is going to be processed.
> > 
> > I would use:
> > 
> >      location /phpmyadmin {
> > 	index index.php;
> >          ...
> > 
> >         location  ~ \.php$ {
> >         ...
> >         }
> >      }
> > 
> > No?
> 
> Yes, inclusive "location  ~ \.php$" is much better
> than configuration
> I suggested.
> 
> However, my configuration should also work, since
> /phpmyadmin/ is
> internally redirected to /phpmyadmin/index.php,
> see for details:
> http://nginx.org/en/docs/http/request_processing.h
> tml#simple_php_site_configuration
> 
> 
> -- 
> Igor Sysoev
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx


I'm trying to explain to you why I can't debug the file, but the forum
says
Possible hack attempt detected. The posted form data was rejected.

I'm having trouble running the "./configure" command, when I try to run
it in terminal

When I uninstalled nginx, somehow apache2 started, while I never
installed apache2
I couldn't uninstall apache, because only the apache2 service was on my
pc, and no installation files.

Maybe this is causing the nginx errors.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200488,200779#msg-200779



From igor at sysoev.ru  Wed May 25 18:42:06 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 18:42:06 +0400
Subject: Configuring phpmyadmin
In-Reply-To: <5fbdb16d9c5bea48f084b81fc00a688d.NginxMailingListEnglish@forum.nginx.org>
References: <20110525140048.GU83054@sysoev.ru>
	<5fbdb16d9c5bea48f084b81fc00a688d.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110525144205.GV83054@sysoev.ru>

> I'm trying to explain to you why I can't debug the file, but the forum
> says
> Possible hack attempt detected. The posted form data was rejected.

The forum is a gateway to mailing list:
http://nginx.org/mailman/listinfo/nginx
You may write there.


-- 
Igor Sysoev


From appa at perusio.net  Wed May 25 19:07:11 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Wed, 25 May 2011 16:07:11 +0100
Subject: redirect subdomain to internal location
In-Reply-To: <30128c32f2b4ce211f2854f9ef451967.NginxMailingListEnglish@forum.nginx.org>
References: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>
	<30128c32f2b4ce211f2854f9ef451967.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87aaealsyo.wl%appa@perusio.net>

On 25 Mai 2011 15h26 WEST, nginx-forum at nginx.us wrote:

> Thanks guys,
>
> I get that now, in my newness to nginx I was overcomplicating
> things. My server_name is now doing what I had hoped to achieve.
>
> If you can bear with me for one more question, how do I get the
> 'client' out of domain.com/client/more/stuff.htm? I want to use the
> 'client' part of the uri to search a redis key/value pair.
>
> I had success with this code for query strings but couldn't work out
> how to get the first part of the uri to be my key?
>
> location / {
> eval_escalate on;
> eval $res {
> set_unescape_uri $key $arg_key;
> redis2_query get $key;
> redis2_pass 127.0.0.1:6379;
> }
> set_by_lua $path "return
> require('redis.parser').parse_reply(ngx.arg[1])" $res;
> proxy_pass http://app_cluster_1/$path;
> proxy_redirect off;
> }

If I'm understanding correctly what you want to do, you'll need a
location with a regex that captures that URI component. E.g.:

location /(?<clientkey>[^/]*)/ {
  (...)
  set_unescape_uri $key $client_key;
  (...)
}

This will match all URIs of the form  /whatever/anything

--- appa




From appa at perusio.net  Wed May 25 19:16:28 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Wed, 25 May 2011 16:16:28 +0100
Subject: redirect subdomain to internal location
In-Reply-To: <87aaealsyo.wl%appa@perusio.net>
References: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>
	<30128c32f2b4ce211f2854f9ef451967.NginxMailingListEnglish@forum.nginx.org>
	<87aaealsyo.wl%appa@perusio.net>
Message-ID: <878vtulsj7.wl%appa@perusio.net>

On 25 Mai 2011 16h07 WEST, appa at perusio.net wrote:

Oops.

> If I'm understanding correctly what you want to do, you'll need a
> location with a regex that captures that URI component. E.g.:
>
> location /(?<clientkey>[^/]*)/ {
> (...)

 set_unescape_uri $key $clientkey;
                       ^^^^^^^^^^
                       Now it's correct.                       
> (...)
> }

--- appa



From appa at perusio.net  Wed May 25 19:20:40 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Wed, 25 May 2011 16:20:40 +0100
Subject: redirect subdomain to internal location
In-Reply-To: <30128c32f2b4ce211f2854f9ef451967.NginxMailingListEnglish@forum.nginx.org>
References: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>
	<30128c32f2b4ce211f2854f9ef451967.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <877h9elsc7.wl%appa@perusio.net>

On 25 Mai 2011 15h26 WEST, nginx-forum at nginx.us wrote:

> Thanks guys,
>
> I get that now, in my newness to nginx I was overcomplicating
> things. My server_name is now doing what I had hoped to achieve.
>
> If you can bear with me for one more question, how do I get the
> 'client' out of domain.com/client/more/stuff.htm? I want to use the
> 'client' part of the uri to search a redis key/value pair.
>
> I had success with this code for query strings but couldn't work out
> how to get the first part of the uri to be my key?
>
> location / {
> eval_escalate on;
> eval $res {
> set_unescape_uri $key $arg_key;
> redis2_query get $key;
> redis2_pass 127.0.0.1:6379;
> }
> set_by_lua $path "return
> require('redis.parser').parse_reply(ngx.arg[1])" $res;
> proxy_pass http://app_cluster_1/$path;
> proxy_redirect off;
> }

If I'm understanding correctly what you want to do, you'll need a
location with a regex that captures that URI component. E.g.:

location ~ /(?<clientkey>[^/]*)/ {
  (...)
  set_unescape_uri $key $clientkey;
  (...)
}

This will match all URIs of the form  /whatever/anything

Forgot the little detail of the ~ for a regex based location. Now it
seems correct.

--- appa




From igor at sysoev.ru  Wed May 25 19:24:53 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Wed, 25 May 2011 19:24:53 +0400
Subject: nginx-1.0.3
Message-ID: <20110525152453.GY83054@sysoev.ru>

Changes with nginx 1.0.3                                         25 May 2011

    *) Feature: the "auth_basic_user_file" directive supports "$apr1", 
       "{PLAIN}", and "{SSHA}" password encryption methods.
       Thanks to Maxim Dounin.

    *) Feature: the "geoip_org" directive and $geoip_org variable.
       Thanks to Alexander Uskov, Arnaud Granal, and Denis F. Latypoff.

    *) Feature: ngx_http_geo_module and ngx_http_geoip_module support IPv4 
       addresses mapped to IPv6 addresses.

    *) Bugfix: a segmentation fault occurred in a worker process during 
       testing IPv4 address mapped to IPv6 address, if access or deny rules 
       were defined only for IPv6; the bug had appeared in 0.8.22.

    *) Bugfix: a cached reponse may be broken if proxy/fastcgi/scgi/ 
       uwsgi_cache_bypass and proxy/fastcgi/scgi/uwsgi_no_cache directive 
       values were different; the bug had appeared in 0.8.46.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Wed May 25 22:14:52 2011
From: nginx-forum at nginx.us (chris.percol)
Date: Wed, 25 May 2011 14:14:52 -0400
Subject: redirect subdomain to internal location
In-Reply-To: <877h9elsc7.wl%appa@perusio.net>
References: <877h9elsc7.wl%appa@perusio.net>
Message-ID: <847822c57ab3d34c8cabbf63d58af5aa.NginxMailingListEnglish@forum.nginx.org>

Ant?nio P. P. Almeida Wrote:
-------------------------------------------------------
> On 25 Mai 2011 15h26 WEST, nginx-forum at nginx.us
> wrote:
> 
> > Thanks guys,
> >
> > I get that now, in my newness to nginx I was
> overcomplicating
> > things. My server_name is now doing what I had
> hoped to achieve.
> >
> > If you can bear with me for one more question,
> how do I get the
> > 'client' out of
> domain.com/client/more/stuff.htm? I want to use
> the
> > 'client' part of the uri to search a redis
> key/value pair.
> >
> > I had success with this code for query strings
> but couldn't work out
> > how to get the first part of the uri to be my
> key?
> >
> > location / {
> > eval_escalate on;
> > eval $res {
> > set_unescape_uri $key $arg_key;
> > redis2_query get $key;
> > redis2_pass 127.0.0.1:6379;
> > }
> > set_by_lua $path "return
> > require('redis.parser').parse_reply(ngx.arg[1])"
> $res;
> > proxy_pass http://app_cluster_1/$path;
> > proxy_redirect off;
> > }
> 
> If I'm understanding correctly what you want to
> do, you'll need a
> location with a regex that captures that URI
> component. E.g.:
> 
> location ~ /(?<clientkey>[^/]*)/ {

Appa,

Thanks again!

My issue is that I don't know the value of clientkey, it could be
several hundred possibilities.

Is there a way to extract the unknown clientkey from
/clientkey/some/more.htm?

Many thanks,

Chris

>   (...)
>   set_unescape_uri $key $clientkey;
>   (...)
> }
> 
> This will match all URIs of the form 
> /whatever/anything
> 
> Forgot the little detail of the ~ for a regex
> based location. Now it
> seems correct.
> 
> --- appa
> 
> 
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200694,200846#msg-200846



From nginx-forum at nginx.us  Wed May 25 22:30:28 2011
From: nginx-forum at nginx.us (the_traveller)
Date: Wed, 25 May 2011 14:30:28 -0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <20110525113708.GJ83054@sysoev.ru>
References: <20110525113708.GJ83054@sysoev.ru>
Message-ID: <c24bafbdb58027c04fde2ee4db29797d.NginxMailingListEnglish@forum.nginx.org>

I know i'm alittle bit annoying

but can you show me how i must write the rule
the path of the script jokes.php is in the public_html dir
and it should talks the args like that 
jokes.php?j=$1

so in Apache  i used  to make the rule like that
RewriteRule ^Joke/.* jokes.php [QSA]

but i alwayes have the sever error on all of my tries

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,23868,200852#msg-200852



From appa at perusio.net  Wed May 25 23:01:30 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Wed, 25 May 2011 20:01:30 +0100
Subject: redirect subdomain to internal location
In-Reply-To: <847822c57ab3d34c8cabbf63d58af5aa.NginxMailingListEnglish@forum.nginx.org>
References: <877h9elsc7.wl%appa@perusio.net>
	<847822c57ab3d34c8cabbf63d58af5aa.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <8762oyli45.wl%appa@perusio.net>

On 25 Mai 2011 19h14 WEST, nginx-forum at nginx.us wrote:


>> location ~ /(?<clientkey>[^/]*)/ {
>
> Appa,
>
> Thanks again!
>
> My issue is that I don't know the value of clientkey, it could be
> several hundred possibilities.
>
> Is there a way to extract the unknown clientkey from
> /clientkey/some/more.htm?
>
> Many thanks,

That's completely generic. It's a named group that captures any char
excepting '/'.

This implicitly creates a variable $clientkey that you can use.

Try it.
--- appa



From nginx-forum at nginx.us  Thu May 26 02:24:01 2011
From: nginx-forum at nginx.us (speedfirst)
Date: Wed, 25 May 2011 18:24:01 -0400
Subject: How to add an arg to the requests to the upstreams
In-Reply-To: <20110525102537.GG83054@sysoev.ru>
References: <20110525102537.GG83054@sysoev.ru>
Message-ID: <91282731f6854359868c392332b842cc.NginxMailingListEnglish@forum.nginx.org>

So how to do that? Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200661,200920#msg-200920



From nginx-forum at nginx.us  Thu May 26 02:28:09 2011
From: nginx-forum at nginx.us (speedfirst)
Date: Wed, 25 May 2011 18:28:09 -0400
Subject: How to add an arg to the requests to the upstreams
In-Reply-To: <91282731f6854359868c392332b842cc.NginxMailingListEnglish@forum.nginx.org>
References: <20110525102537.GG83054@sysoev.ru>
	<91282731f6854359868c392332b842cc.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <0b232b6e659e6e32dda813e1f9691549.NginxMailingListEnglish@forum.nginx.org>

sorry, ignore the last pst.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200661,200924#msg-200924



From nginx-forum at nginx.us  Thu May 26 02:41:46 2011
From: nginx-forum at nginx.us (speedfirst)
Date: Wed, 25 May 2011 18:41:46 -0400
Subject: How to make a redirect via programming?
Message-ID: <f13827a6a3e490f7e3e1295b72275a5f.NginxMailingListEnglish@forum.nginx.org>

Hey, I'm making a custom module and need to send back the redirect
response to the client. It should behave like the "sendRedirect" in java
servlet programming. So which function is that API?

Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200929,200929#msg-200929



From nginx-forum at nginx.us  Thu May 26 02:57:25 2011
From: nginx-forum at nginx.us (phil1886)
Date: Wed, 25 May 2011 18:57:25 -0400
Subject: Location directive disaster
Message-ID: <63308ba80e89b53b19c992ed504da4a6.NginxMailingListEnglish@forum.nginx.org>

Hi everyone

I have moved from Apache2 to Nginx. I am finding it difficult working
out what is going on with the location directive. I've looked at various
tutorials and blogs but I'm still encountering problems.

Basically I have a new website that I'm hosting off a directory of the
current one. Lets call it /newwebsite. To prevent people getting to it,
I've put http-basic authentication around it:

                ## Default Location
                location / {
                        index   index.php       index.html;
                }

                location /newwebsite {
                        index   index.php       index.html;
                        auth_basic              "Unauthorised use
prohibited";
                        auth_basic_user_file   
/home/username/www.website.com/.htpasswd;
                }

There's an admin section as well. I will need http authentication on
this, even when it goes live, to stop people gaining access to the admin
function:

                location /newwebsite/admin {
                        index login.php;
                        auth_basic              "Unauthorised use
prohibited";
                        auth_basic_user_file   
/home/username/www.website.com/.htpasswd;

                }

Of course I have the catch-all for all the php scripts:

                location ~ \.php$ {
                        include /etc/nginx/fastcgi_params;
                        fastcgi_pass    127.0.0.1:9000;
                        fastcgi_index   index.php;
                        fastcgi_param   SCRIPT_FILENAME        
$document_root$fastcgi_script_name;
                        fastcgi_param   PATH_INFO      
$fastcgi_script_name;
                }

So:

When I visit http://domain.com/newsite/admin I get prompted for a http
basic-auth username and password - as expected. This works.

When I visit http://domain.com/newsite/admin/admin.php I am straight in
without http basic-auth. This is not so good..!!

So perhaps I'm doing this all wrong?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200935,200935#msg-200935



From nginx-forum at nginx.us  Thu May 26 03:14:59 2011
From: nginx-forum at nginx.us (speedfirst)
Date: Wed, 25 May 2011 19:14:59 -0400
Subject: How to make a redirect via programming?
In-Reply-To: <f13827a6a3e490f7e3e1295b72275a5f.NginxMailingListEnglish@forum.nginx.org>
References: <f13827a6a3e490f7e3e1295b72275a5f.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <15238faaa7505eea2c561c97cdb9dade.NginxMailingListEnglish@forum.nginx.org>

is this OK?

r->headers_out.status = NGX_HTTP_REDIRECT;
r->headers_out.content_length_n = 0;
... // the code set "Location" header

ngx_http_send_header(r);

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200929,200937#msg-200937



From tomlove at gmail.com  Thu May 26 03:15:34 2011
From: tomlove at gmail.com (Thomas Love)
Date: Thu, 26 May 2011 01:15:34 +0200
Subject: Location directive disaster
In-Reply-To: <63308ba80e89b53b19c992ed504da4a6.NginxMailingListEnglish@forum.nginx.org>
References: <63308ba80e89b53b19c992ed504da4a6.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTikUdwXh0kY=h37co5gYO-QNmSzyTg@mail.gmail.com>

On 26 May 2011 00:57, phil1886 <nginx-forum at nginx.us> wrote:

> Hi everyone
>
>
Hi.


>                location /newwebsite/admin {
>                        index login.php;
>                        auth_basic              "Unauthorised use
> prohibited";
>                        auth_basic_user_file
> /home/username/www.website.com/.htpasswd;
>
>                }
>
> Of course I have the catch-all for all the php scripts:
>
>                location ~ \.php$ {
>                        include /etc/nginx/fastcgi_params;
>                        fastcgi_pass    127.0.0.1:9000;
>                        fastcgi_index   index.php;
>                        fastcgi_param   SCRIPT_FILENAME
> $document_root$fastcgi_script_name;
>                        fastcgi_param   PATH_INFO
> $fastcgi_script_name;
>                }
>
>
> When I visit http://domain.com/newsite/admin/admin.php I am straight in
> without http basic-auth. This is not so good..!!
>
>


You have to put a copy of the ~ \.php$ location block inside the protected
location block (nested), otherwise the php block takes precedence and the
location parser does not match anything else.

I had exactly the same problem: http://forum.nginx.org/read.php?2,189220

Thomas
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110526/542c8ee3/attachment.html>

From nginx-forum at nginx.us  Thu May 26 03:41:45 2011
From: nginx-forum at nginx.us (phil1886)
Date: Wed, 25 May 2011 19:41:45 -0400
Subject: Location directive disaster
In-Reply-To: <63308ba80e89b53b19c992ed504da4a6.NginxMailingListEnglish@forum.nginx.org>
References: <63308ba80e89b53b19c992ed504da4a6.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <272c8a57b15ed7b85fff2242876dc9ad.NginxMailingListEnglish@forum.nginx.org>

Thank you!!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200935,200942#msg-200942



From nginx-forum at nginx.us  Thu May 26 04:39:15 2011
From: nginx-forum at nginx.us (wyatt)
Date: Wed, 25 May 2011 20:39:15 -0400
Subject: NGINX 1.0.3 compile error: undefined reference to `ngx_libc_crypt'
Message-ID: <095bd58064177a9e78cb9129ade3170a.NginxMailingListEnglish@forum.nginx.org>

Hey Guys,

I've found a compile error for NGINX 1.0.3 that didn't exist for any of
the previous versions (same configuration, etc.) Here's the
"./configure" arguments I used:


./configure --sbin-path=/usr/local/sbin --with-http_ssl_module
--with-ipv6 --without-mail_pop3_module --without-mail_imap_module
--without-mail_smtp_module --without-http_autoindex_module
--without-http_ssi_module --without-http_auth_basic_module
--without-http_browser_module --without-http_empty_gif_module
--without-http_geo_module --without-http_limit_req_module
--without-http_limit_zone_module --without-http_map_module
--without-http_memcached_module --without-http_proxy_module
--without-http_referer_module --without-http_scgi_module
--without-http_split_clients_module
--without-http_upstream_ip_hash_module --without-http_userid_module
--without-http_uwsgi_module --with-cc-opt=-O2


Then I ran "make"

After a little while, this is the ouput:



gcc -o objs/nginx \
	objs/src/core/nginx.o \
[.....]
	objs/ngx_modules.o \
	-lpcre -lssl -lcrypto -ldl -lz
objs/src/core/ngx_crypt.o: In function `ngx_crypt':
/home/wyatt/sources/nginx-1.0.3/src/core/ngx_crypt.c:48: undefined
reference to `ngx_libc_crypt'
collect2: ld returned 1 exit status
make[1]: *** [objs/nginx] Error 1
make[1]: Leaving directory `/home/wyatt/sources/nginx-1.0.3'
make: *** [build] Error 2




Am I doing something wrong, or is the problem somewhere in the
./configure or make scripts?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200948,200948#msg-200948



From kworthington at gmail.com  Thu May 26 05:17:32 2011
From: kworthington at gmail.com (Kevin Worthington)
Date: Wed, 25 May 2011 21:17:32 -0400
Subject: Nginx-1.0.3 for Windows 32-bit and 64-bit now available
Message-ID: <BANLkTi=VnGTnaYeT7XEcFBoKxj0Sa-+rcQ@mail.gmail.com>

For those interested, Cygwin based versions of Nginx 1.0.3 for Windows
(both 32-bit and 64-bit) builds are available here:
http://www.kevinworthington.com/nginx-for-windows/

Blog post/announcement:
http://www.kevinworthington.com/nginx-1-0-3-for-windows/ (comments and
feedback welcome)

Cygwin is NOT required on your system to run this version; all needed
files are included.

Please note that I have been providing Cygwin based builds since
before official Windows builds were made available. It's mostly to
support folks already using the builds.

Thank you,
Kevin
--
Kevin Worthington
http://www.kevinworthington.com/


From wendal1985 at gmail.com  Thu May 26 06:09:27 2011
From: wendal1985 at gmail.com (Wendal Chen)
Date: Thu, 26 May 2011 10:09:27 +0800
Subject: nginx-1.0.3
In-Reply-To: <20110525152453.GY83054@sysoev.ru>
References: <20110525152453.GY83054@sysoev.ru>
Message-ID: <BANLkTikJKkfA3D_aGzeZjKFUd3rUgtbMTA@mail.gmail.com>

It has a bug!

Build with :
./configure --without-http_auth_basic_module && make

objs/src/core/ngx_crypt.o: In function `ngx_crypt':
ngx_crypt.c:(.text+0x7b): undefined reference to `ngx_libc_crypt'
collect2: ld returned 1 exit status
make[1]: *** [objs/nginx] Error 1
make[1]: Leaving directory `/root/nginx-1.0.3'
make: *** [build] Error 2


But ok when:
./configure && make

2011/5/25 Igor Sysoev <igor at sysoev.ru>

> Changes with nginx 1.0.3                                         25 May
> 2011
>
>    *) Feature: the "auth_basic_user_file" directive supports "$apr1",
>       "{PLAIN}", and "{SSHA}" password encryption methods.
>       Thanks to Maxim Dounin.
>
>    *) Feature: the "geoip_org" directive and $geoip_org variable.
>       Thanks to Alexander Uskov, Arnaud Granal, and Denis F. Latypoff.
>
>    *) Feature: ngx_http_geo_module and ngx_http_geoip_module support IPv4
>       addresses mapped to IPv6 addresses.
>
>    *) Bugfix: a segmentation fault occurred in a worker process during
>       testing IPv4 address mapped to IPv6 address, if access or deny rules
>       were defined only for IPv6; the bug had appeared in 0.8.22.
>
>    *) Bugfix: a cached reponse may be broken if proxy/fastcgi/scgi/
>       uwsgi_cache_bypass and proxy/fastcgi/scgi/uwsgi_no_cache directive
>       values were different; the bug had appeared in 0.8.46.
>
>
> --
> Igor Sysoev
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>



-- 
Wendal Chen
<http://t.sina.com.cn/1068040557?s=6uyXnP>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110526/55e745b8/attachment.html>

From wendal1985 at gmail.com  Thu May 26 06:13:36 2011
From: wendal1985 at gmail.com (Wendal Chen)
Date: Thu, 26 May 2011 10:13:36 +0800
Subject: NGINX 1.0.3 compile error: undefined reference to `ngx_libc_crypt'
In-Reply-To: <095bd58064177a9e78cb9129ade3170a.NginxMailingListEnglish@forum.nginx.org>
References: <095bd58064177a9e78cb9129ade3170a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTi=ZZ-_AogiqV6Yb01bLitHz7wLeAw@mail.gmail.com>

remove :      --without-http_auth_basic_module

It will be ok.

I think it is a Bug!!


2011/5/26 wyatt <nginx-forum at nginx.us>

> Hey Guys,
>
> I've found a compile error for NGINX 1.0.3 that didn't exist for any of
> the previous versions (same configuration, etc.) Here's the
> "./configure" arguments I used:
>
>
> ./configure --sbin-path=/usr/local/sbin --with-http_ssl_module
> --with-ipv6 --without-mail_pop3_module --without-mail_imap_module
> --without-mail_smtp_module --without-http_autoindex_module
> --without-http_ssi_module --without-http_auth_basic_module
> --without-http_browser_module --without-http_empty_gif_module
> --without-http_geo_module --without-http_limit_req_module
> --without-http_limit_zone_module --without-http_map_module
> --without-http_memcached_module --without-http_proxy_module
> --without-http_referer_module --without-http_scgi_module
> --without-http_split_clients_module
> --without-http_upstream_ip_hash_module --without-http_userid_module
> --without-http_uwsgi_module --with-cc-opt=-O2
>
>
> Then I ran "make"
>
> After a little while, this is the ouput:
>
>
>
> gcc -o objs/nginx \
>        objs/src/core/nginx.o \
> [.....]
>        objs/ngx_modules.o \
>        -lpcre -lssl -lcrypto -ldl -lz
> objs/src/core/ngx_crypt.o: In function `ngx_crypt':
> /home/wyatt/sources/nginx-1.0.3/src/core/ngx_crypt.c:48: undefined
> reference to `ngx_libc_crypt'
> collect2: ld returned 1 exit status
> make[1]: *** [objs/nginx] Error 1
> make[1]: Leaving directory `/home/wyatt/sources/nginx-1.0.3'
> make: *** [build] Error 2
>
>
>
>
> Am I doing something wrong, or is the problem somewhere in the
> ./configure or make scripts?
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,200948,200948#msg-200948
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>



-- 
Wendal Chen
<http://t.sina.com.cn/1068040557?s=6uyXnP>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110526/01a08b0f/attachment.html>

From wendal1985 at gmail.com  Thu May 26 06:47:48 2011
From: wendal1985 at gmail.com (Wendal Chen)
Date: Thu, 26 May 2011 10:47:48 +0800
Subject: nginx-1.0.3
In-Reply-To: <BANLkTikJKkfA3D_aGzeZjKFUd3rUgtbMTA@mail.gmail.com>
References: <20110525152453.GY83054@sysoev.ru>
	<BANLkTikJKkfA3D_aGzeZjKFUd3rUgtbMTA@mail.gmail.com>
Message-ID: <BANLkTim-JkmtZNmVVuhypOJSfw3gysoCcg@mail.gmail.com>

I think there is the point:

in ngx_crypt.c
ngx_int_t
ngx_crypt(ngx_pool_t *pool, u_char *key, u_char *salt, u_char **encrypted)
{
    if (ngx_strncmp(salt, "$apr1$", sizeof("$apr1$") - 1) == 0) {
        return ngx_crypt_apr1(pool, key, salt, encrypted);
    } else if (ngx_strncmp(salt, "{PLAIN}", sizeof("{PLAIN}") - 1) == 0) {
        return ngx_crypt_plain(pool, key, salt, encrypted);

#if (NGX_HAVE_SHA1)
    } else if (ngx_strncmp(salt, "{SSHA}", sizeof("{SSHA}") - 1) == 0) {
        return ngx_crypt_ssha(pool, key, salt, encrypted);
#endif
    }

    /* fallback to libc crypt() */
    return ngx_libc_crypt(pool, key, salt, encrypted);
}

and there is no libc crypt in my system.

So , I think is a bug .

2011/5/26 Wendal Chen <wendal1985 at gmail.com>

> It has a bug!
>
> Build with :
> ./configure --without-http_auth_basic_module && make
>
> objs/src/core/ngx_crypt.o: In function `ngx_crypt':
> ngx_crypt.c:(.text+0x7b): undefined reference to `ngx_libc_crypt'
> collect2: ld returned 1 exit status
> make[1]: *** [objs/nginx] Error 1
> make[1]: Leaving directory `/root/nginx-1.0.3'
> make: *** [build] Error 2
>
>
> But ok when:
> ./configure && make
>
>
> 2011/5/25 Igor Sysoev <igor at sysoev.ru>
>
>> Changes with nginx 1.0.3                                         25 May
>> 2011
>>
>>    *) Feature: the "auth_basic_user_file" directive supports "$apr1",
>>       "{PLAIN}", and "{SSHA}" password encryption methods.
>>       Thanks to Maxim Dounin.
>>
>>    *) Feature: the "geoip_org" directive and $geoip_org variable.
>>       Thanks to Alexander Uskov, Arnaud Granal, and Denis F. Latypoff.
>>
>>    *) Feature: ngx_http_geo_module and ngx_http_geoip_module support IPv4
>>       addresses mapped to IPv6 addresses.
>>
>>    *) Bugfix: a segmentation fault occurred in a worker process during
>>       testing IPv4 address mapped to IPv6 address, if access or deny rules
>>       were defined only for IPv6; the bug had appeared in 0.8.22.
>>
>>    *) Bugfix: a cached reponse may be broken if proxy/fastcgi/scgi/
>>       uwsgi_cache_bypass and proxy/fastcgi/scgi/uwsgi_no_cache directive
>>       values were different; the bug had appeared in 0.8.46.
>>
>>
>> --
>> Igor Sysoev
>>
>> _______________________________________________
>> nginx mailing list
>> nginx at nginx.org
>> http://nginx.org/mailman/listinfo/nginx
>>
>
>
>
> --
> Wendal Chen
> <http://t.sina.com.cn/1068040557?s=6uyXnP>
>
>


-- 
Wendal Chen
<http://t.sina.com.cn/1068040557?s=6uyXnP>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110526/357cfebb/attachment.html>

From josh at joshparker.us  Thu May 26 07:03:52 2011
From: josh at joshparker.us (=?utf-8?Q?Parker=2C_Joshua?=)
Date: Wed, 25 May 2011 23:03:52 -0400
Subject: Error with http_perl_module
Message-ID: <7104EC025B8F4C558C84166030307E0C@joshparker.us>

I am trying to upgrade to 1.0.2, but I am getting an install error. Any help is greatly appreciated.

make[2]: Leaving directory `/usr/local/src/nginx-1.0.2/objs/src/http/modules/perl'
rm -rf objs/install_perl
gcc -o objs/nginx \
objs/src/core/nginx.o \
objs/src/core/ngx_log.o \
objs/src/core/ngx_palloc.o \
objs/src/core/ngx_array.o \
objs/src/core/ngx_list.o \
objs/src/core/ngx_hash.o \
objs/src/core/ngx_buf.o \
objs/src/core/ngx_queue.o \
objs/src/core/ngx_output_chain.o \
objs/src/core/ngx_string.o \
objs/src/core/ngx_parse.o \
objs/src/core/ngx_inet.o \
objs/src/core/ngx_file.o \
objs/src/core/ngx_crc32.o \
objs/src/core/ngx_murmurhash.o \
objs/src/core/ngx_md5.o \
objs/src/core/ngx_rbtree.o \
objs/src/core/ngx_radix_tree.o \
objs/src/core/ngx_slab.o \
objs/src/core/ngx_times.o \
objs/src/core/ngx_shmtx.o \
objs/src/core/ngx_connection.o \
objs/src/core/ngx_cycle.o \
objs/src/core/ngx_spinlock.o \
objs/src/core/ngx_cpuinfo.o \
objs/src/core/ngx_conf_file.o \
objs/src/core/ngx_resolver.o \
objs/src/core/ngx_open_file_cache.o \
objs/src/event/ngx_event.o \
objs/src/event/ngx_event_timer.o \
objs/src/event/ngx_event_posted.o \
objs/src/event/ngx_event_busy_lock.o \
objs/src/event/ngx_event_accept.o \
objs/src/event/ngx_event_connect.o \
objs/src/event/ngx_event_pipe.o \
objs/src/os/unix/ngx_time.o \
objs/src/os/unix/ngx_errno.o \
objs/src/os/unix/ngx_alloc.o \
objs/src/os/unix/ngx_files.o \
objs/src/os/unix/ngx_socket.o \
objs/src/os/unix/ngx_recv.o \
objs/src/os/unix/ngx_readv_chain.o \
objs/src/os/unix/ngx_udp_recv.o \
objs/src/os/unix/ngx_send.o \
objs/src/os/unix/ngx_writev_chain.o \
objs/src/os/unix/ngx_channel.o \
objs/src/os/unix/ngx_shmem.o \
objs/src/os/unix/ngx_process.o \
objs/src/os/unix/ngx_daemon.o \
objs/src/os/unix/ngx_setproctitle.o \
objs/src/os/unix/ngx_posix_init.o \
objs/src/os/unix/ngx_user.o \
objs/src/os/unix/ngx_process_cycle.o \
objs/src/os/unix/ngx_linux_init.o \
objs/src/event/modules/ngx_epoll_module.o \
objs/src/os/unix/ngx_linux_sendfile_chain.o \
objs/src/event/ngx_event_openssl.o \
objs/src/core/ngx_regex.o \
objs/src/http/ngx_http.o \
objs/src/http/ngx_http_core_module.o \
objs/src/http/ngx_http_special_response.o \
objs/src/http/ngx_http_request.o \
objs/src/http/ngx_http_parse.o \
objs/src/http/ngx_http_header_filter_module.o \
objs/src/http/ngx_http_write_filter_module.o \
objs/src/http/ngx_http_copy_filter_module.o \
objs/src/http/modules/ngx_http_log_module.o \
objs/src/http/ngx_http_request_body.o \
objs/src/http/ngx_http_variables.o \
objs/src/http/ngx_http_script.o \
objs/src/http/ngx_http_upstream.o \
objs/src/http/ngx_http_upstream_round_robin.o \
objs/src/http/ngx_http_parse_time.o \
objs/src/http/modules/ngx_http_static_module.o \
objs/src/http/modules/ngx_http_index_module.o \
objs/src/http/modules/ngx_http_chunked_filter_module.o \
objs/src/http/modules/ngx_http_range_filter_module.o \
objs/src/http/modules/ngx_http_headers_filter_module.o \
objs/src/http/modules/ngx_http_not_modified_filter_module.o \
objs/src/http/ngx_http_busy_lock.o \
objs/src/http/ngx_http_file_cache.o \
objs/src/http/modules/ngx_http_gzip_filter_module.o \
objs/src/http/ngx_http_postpone_filter_module.o \
objs/src/http/modules/ngx_http_ssi_filter_module.o \
objs/src/http/modules/ngx_http_charset_filter_module.o \
objs/src/http/modules/ngx_http_userid_filter_module.o \
objs/src/http/modules/ngx_http_autoindex_module.o \
objs/src/http/modules/ngx_http_auth_basic_module.o \
objs/src/http/modules/ngx_http_access_module.o \
objs/src/http/modules/ngx_http_limit_zone_module.o \
objs/src/http/modules/ngx_http_limit_req_module.o \
objs/src/http/modules/ngx_http_geo_module.o \
objs/src/http/modules/ngx_http_map_module.o \
objs/src/http/modules/ngx_http_split_clients_module.o \
objs/src/http/modules/ngx_http_referer_module.o \
objs/src/http/modules/ngx_http_rewrite_module.o \
objs/src/http/modules/ngx_http_ssl_module.o \
objs/src/http/modules/ngx_http_proxy_module.o \
objs/src/http/modules/ngx_http_fastcgi_module.o \
objs/src/http/modules/ngx_http_uwsgi_module.o \
objs/src/http/modules/ngx_http_scgi_module.o \
objs/src/http/modules/perl/ngx_http_perl_module.o \
objs/src/http/modules/ngx_http_memcached_module.o \
objs/src/http/modules/ngx_http_empty_gif_module.o \
objs/src/http/modules/ngx_http_browser_module.o \
objs/src/http/modules/ngx_http_flv_module.o \
objs/src/http/modules/ngx_http_upstream_ip_hash_module.o \
objs/src/http/modules/ngx_http_stub_status_module.o \
objs/src/mail/ngx_mail.o \
objs/src/mail/ngx_mail_core_module.o \
objs/src/mail/ngx_mail_handler.o \
objs/src/mail/ngx_mail_parse.o \
objs/src/mail/ngx_mail_ssl_module.o \
objs/src/mail/ngx_mail_pop3_module.o \
objs/src/mail/ngx_mail_pop3_handler.o \
objs/src/mail/ngx_mail_imap_module.o \
objs/src/mail/ngx_mail_imap_handler.o \
objs/src/mail/ngx_mail_smtp_module.o \
objs/src/mail/ngx_mail_smtp_handler.o \
objs/src/mail/ngx_mail_auth_http_module.o \
objs/src/mail/ngx_mail_proxy_module.o \
objs/ngx_modules.o \
-lcrypt -lpcre -lssl -lcrypto -ldl -lz \
-Wl,-E -fstack-protector -L/usr/local/lib -L/usr/lib/perl/5.10/CORE -ldl -lm -lpthread -lc -lcrypt
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_init_interpreter':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:516: undefined reference to `Perl_sys_init'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_create_interpreter':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:560: undefined reference to `perl_alloc'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:569: undefined reference to `PL_thr_key'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:569: undefined reference to `Perl_croak_nocontext'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:571: undefined reference to `perl_construct'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:598: undefined reference to `perl_parse'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:605: undefined reference to `Perl_get_sv'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:606: undefined reference to `Perl_sv_2pv_flags'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_run_requires':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:648: undefined reference to `Perl_require_pv'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:650: undefined reference to `Perl_sv_2bool'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:652: undefined reference to `Perl_sv_2pv_flags'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_create_interpreter':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:625: undefined reference to `perl_destruct'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:627: undefined reference to `perl_free'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_xs_init':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:162: undefined reference to `boot_DynaLoader'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:162: undefined reference to `Perl_newXS'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:164: undefined reference to `Perl_gv_stashpv'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_cleanup_perl':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:828: undefined reference to `PL_thr_key'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:828: undefined reference to `Perl_croak_nocontext'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:830: undefined reference to `perl_destruct'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:832: undefined reference to `perl_free'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:837: undefined reference to `Perl_sys_term'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_eval_anon_sub':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:778: undefined reference to `Perl_eval_pv'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_set':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:1007: undefined reference to `PL_thr_key'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:1007: undefined reference to `Perl_croak_nocontext'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:1018: undefined reference to `Perl_newSVpvn'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:936: undefined reference to `PL_thr_key'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:936: undefined reference to `Perl_croak_nocontext'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:947: undefined reference to `Perl_newSVpvn'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_call_handler':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:683: undefined reference to `Perl_push_scope'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:684: undefined reference to `Perl_save_int'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:686: undefined reference to `Perl_markstack_grow'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:688: undefined reference to `Perl_newSViv'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:688: undefined reference to `Perl_newRV_noinc'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:688: undefined reference to `Perl_sv_bless'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:688: undefined reference to `Perl_sv_2mortal'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:689: undefined reference to `Perl_stack_grow'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:692: undefined reference to `Perl_stack_grow'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:695: undefined reference to `Perl_sv_2mortal'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:703: undefined reference to `Perl_call_sv'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:709: undefined reference to `Perl_sv_2iv_flags'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:715: undefined reference to `Perl_sv_2pv_flags'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:729: undefined reference to `Perl_free_tmps'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:730: undefined reference to `Perl_pop_scope'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:734: undefined reference to `Perl_sv_2bool'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:736: undefined reference to `Perl_sv_2pv_flags'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_variable':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:323: undefined reference to `PL_thr_key'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:323: undefined reference to `Perl_croak_nocontext'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_ssi':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:388: undefined reference to `PL_thr_key'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:388: undefined reference to `Perl_croak_nocontext'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:408: undefined reference to `Perl_newSVpvn'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:419: undefined reference to `Perl_sv_free2'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:419: undefined reference to `Perl_sv_free'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:426: undefined reference to `Perl_newSVpvn'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:436: undefined reference to `Perl_sv_free2'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:436: undefined reference to `Perl_sv_free'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_init_worker':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:1039: undefined reference to `PL_thr_key'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:1039: undefined reference to `Perl_croak_nocontext'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:1043: undefined reference to `Perl_gv_fetchpv'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:1043: undefined reference to `Perl_sv_setiv'
objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_handle_request':
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:208: undefined reference to `PL_thr_key'
/usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:208: undefined reference to `Perl_croak_nocontext'
collect2: ld returned 1 exit status
make[1]: *** [objs/nginx] Error 1
make[1]: Leaving directory `/usr/local/src/nginx-1.0.2'
make: *** [install] Error 2

-- 
Parker, Joshua
Sent with Sparrow (http://www.sparrowmailapp.com/?sig)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110525/ba013a71/attachment-0001.html>

From nginx-forum at nginx.us  Thu May 26 07:50:47 2011
From: nginx-forum at nginx.us (wyatt)
Date: Wed, 25 May 2011 23:50:47 -0400
Subject: NGINX 1.0.3 compile error: undefined reference to `ngx_libc_crypt'
In-Reply-To: <BANLkTi=ZZ-_AogiqV6Yb01bLitHz7wLeAw@mail.gmail.com>
References: <BANLkTi=ZZ-_AogiqV6Yb01bLitHz7wLeAw@mail.gmail.com>
Message-ID: <96abe5c394040f7411fe32d5376c044f.NginxMailingListEnglish@forum.nginx.org>

Thanks, that worked.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200948,201006#msg-201006



From nginx-forum at nginx.us  Thu May 26 08:13:51 2011
From: nginx-forum at nginx.us (agung)
Date: Thu, 26 May 2011 00:13:51 -0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <c0c6df7135604bcee44a098e5b896a30@ruby-forum.com>
References: <c0c6df7135604bcee44a098e5b896a30@ruby-forum.com>
Message-ID: <0425005b61c87f8219125423c2e06c4c.NginxMailingListEnglish@forum.nginx.org>

I still do not understand the previous explanation

Can you help me to convert my htaccess to nginx...
 

RewriteEngine on
RewriteRule ^$ index.php [L]
RewriteCond $1
!^(index\.php|info\.php|dotpro|poploud|users|upload_pic|static|oauth|sharefb|dotproject|error|facebook|blog|system/application/views/(apps-images|css-libraries|js-libraries|swf-library|video-library)|AuthSub\.php|robots\.txt|favicon\.ico|insecure\.php)
RewriteRule ^(.*)$ index.php/$1 [L]

thank's

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,23868,201013#msg-201013



From igor at sysoev.ru  Thu May 26 09:31:13 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Thu, 26 May 2011 09:31:13 +0400
Subject: NGINX 1.0.3 compile error: undefined reference to `ngx_libc_crypt'
In-Reply-To: <095bd58064177a9e78cb9129ade3170a.NginxMailingListEnglish@forum.nginx.org>
References: <095bd58064177a9e78cb9129ade3170a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110526053113.GC86377@sysoev.ru>

On Wed, May 25, 2011 at 08:39:15PM -0400, wyatt wrote:
> Hey Guys,
> 
> I've found a compile error for NGINX 1.0.3 that didn't exist for any of
> the previous versions (same configuration, etc.) Here's the
> "./configure" arguments I used:
> 
> 
> ./configure --sbin-path=/usr/local/sbin --with-http_ssl_module
> --with-ipv6 --without-mail_pop3_module --without-mail_imap_module
> --without-mail_smtp_module --without-http_autoindex_module
> --without-http_ssi_module --without-http_auth_basic_module
> --without-http_browser_module --without-http_empty_gif_module
> --without-http_geo_module --without-http_limit_req_module
> --without-http_limit_zone_module --without-http_map_module
> --without-http_memcached_module --without-http_proxy_module
> --without-http_referer_module --without-http_scgi_module
> --without-http_split_clients_module
> --without-http_upstream_ip_hash_module --without-http_userid_module
> --without-http_uwsgi_module --with-cc-opt=-O2
> 
> 
> Then I ran "make"
> 
> After a little while, this is the ouput:
> 
> 
> 
> gcc -o objs/nginx \
> 	objs/src/core/nginx.o \
> [.....]
> 	objs/ngx_modules.o \
> 	-lpcre -lssl -lcrypto -ldl -lz
> objs/src/core/ngx_crypt.o: In function `ngx_crypt':
> /home/wyatt/sources/nginx-1.0.3/src/core/ngx_crypt.c:48: undefined
> reference to `ngx_libc_crypt'
> collect2: ld returned 1 exit status
> make[1]: *** [objs/nginx] Error 1
> make[1]: Leaving directory `/home/wyatt/sources/nginx-1.0.3'
> make: *** [build] Error 2
> 
> 
> 
> 
> Am I doing something wrong, or is the problem somewhere in the
> ./configure or make scripts?

The attached patch fixes the issue.


-- 
Igor Sysoev
-------------- next part --------------
Index: src/core/ngx_crypt.c
===================================================================
--- src/core/ngx_crypt.c	(revision 3925)
+++ src/core/ngx_crypt.c	(working copy)
@@ -12,6 +12,8 @@
 #endif
 
 
+#if (NGX_CRYPT)
+
 static ngx_int_t ngx_crypt_apr1(ngx_pool_t *pool, u_char *key, u_char *salt,
     u_char **encrypted);
 static ngx_int_t ngx_crypt_plain(ngx_pool_t *pool, u_char *key, u_char *salt,
@@ -232,3 +234,5 @@
 }
 
 #endif /* NGX_HAVE_SHA1 */
+
+#endif /* NGX_CRYPT */

From igor at sysoev.ru  Thu May 26 11:05:03 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Thu, 26 May 2011 11:05:03 +0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <c24bafbdb58027c04fde2ee4db29797d.NginxMailingListEnglish@forum.nginx.org>
References: <20110525113708.GJ83054@sysoev.ru>
	<c24bafbdb58027c04fde2ee4db29797d.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110526070503.GG86377@sysoev.ru>

On Wed, May 25, 2011 at 02:30:28PM -0400, the_traveller wrote:
> I know i'm alittle bit annoying
> 
> but can you show me how i must write the rule
> the path of the script jokes.php is in the public_html dir
> and it should talks the args like that 
> jokes.php?j=$1

Where does this public_html directory reside ?
/usr/local/www, /home/user/, or anything else ?


-- 
Igor Sysoev


From igor at sysoev.ru  Thu May 26 11:40:48 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Thu, 26 May 2011 11:40:48 +0400
Subject: Error with http_perl_module
In-Reply-To: <7104EC025B8F4C558C84166030307E0C@joshparker.us>
References: <7104EC025B8F4C558C84166030307E0C@joshparker.us>
Message-ID: <20110526074048.GI86377@sysoev.ru>

On Wed, May 25, 2011 at 11:03:52PM -0400, Parker, Joshua wrote:
> I am trying to upgrade to 1.0.2, but I am getting an install error. Any help is greatly appreciated.

> -lcrypt -lpcre -lssl -lcrypto -ldl -lz \
> -Wl,-E -fstack-protector -L/usr/local/lib -L/usr/lib/perl/5.10/CORE -ldl -lm -lpthread -lc -lcrypt
> objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_init_interpreter':
> /usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:516: undefined reference to `Perl_sys_init'

What does "perl -MExtUtils::Embed -e ldopts" show ?


-- 
Igor Sysoev


From mdounin at mdounin.ru  Thu May 26 11:47:29 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 26 May 2011 11:47:29 +0400
Subject: nginx-1.0.3
In-Reply-To: <BANLkTikJKkfA3D_aGzeZjKFUd3rUgtbMTA@mail.gmail.com>
References: <20110525152453.GY83054@sysoev.ru>
	<BANLkTikJKkfA3D_aGzeZjKFUd3rUgtbMTA@mail.gmail.com>
Message-ID: <20110526074729.GC42265@mdounin.ru>

Hello!

On Thu, May 26, 2011 at 10:09:27AM +0800, Wendal Chen wrote:

> It has a bug!
> 
> Build with :
> ./configure --without-http_auth_basic_module && make
> 
> objs/src/core/ngx_crypt.o: In function `ngx_crypt':
> ngx_crypt.c:(.text+0x7b): undefined reference to `ngx_libc_crypt'
> collect2: ld returned 1 exit status
> make[1]: *** [objs/nginx] Error 1
> make[1]: Leaving directory `/root/nginx-1.0.3'
> make: *** [build] Error 2
> 
> 
> But ok when:
> ./configure && make

Thank you for your report.  The following patch unbreaks build 
without auth basic module:

diff --git a/src/core/ngx_crypt.c b/src/core/ngx_crypt.c
--- a/src/core/ngx_crypt.c
+++ b/src/core/ngx_crypt.c
@@ -11,6 +11,7 @@
 #include <ngx_sha1.h>
 #endif

+#if (NGX_CRYPT)

 static ngx_int_t ngx_crypt_apr1(ngx_pool_t *pool, u_char *key, u_char *salt,
     u_char **encrypted);
@@ -232,3 +233,5 @@ ngx_crypt_ssha(ngx_pool_t *pool, u_char
 }

 #endif /* NGX_HAVE_SHA1 */
+
+#endif /* NGX_CRYPT */


Maxim Dounin


From mdounin at mdounin.ru  Thu May 26 12:03:26 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 26 May 2011 12:03:26 +0400
Subject: How to make a redirect via programming?
In-Reply-To: <f13827a6a3e490f7e3e1295b72275a5f.NginxMailingListEnglish@forum.nginx.org>
References: <f13827a6a3e490f7e3e1295b72275a5f.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110526080326.GD42265@mdounin.ru>

Hello!

On Wed, May 25, 2011 at 06:41:46PM -0400, speedfirst wrote:

> Hey, I'm making a custom module and need to send back the redirect
> response to the client. It should behave like the "sendRedirect" in java
> servlet programming. So which function is that API?
> 
> Thanks.

    r->headers_out.location = ngx_palloc(r->pool, sizeof(ngx_table_elt_t));
    if (r->headers_out.location == NULL) {
        return NGX_HTTP_INTERNAL_SERVER_ERROR;
    }

    r->headers_out.location->value.len = sizeof("http://example.com/") - 1;
    r->headers_out.location->value.data = "http://example.com/";

    return NGX_HTTP_MOVED_PERMANENTLY;

Maxim Dounin


From Richard.Kearsley at m247.com  Thu May 26 12:06:37 2011
From: Richard.Kearsley at m247.com (Richard Kearsley)
Date: Thu, 26 May 2011 08:06:37 +0000
Subject: core dumps
Message-ID: <EFC5A19DBF3A2E489904733B467D990B131183@office-server2.m247.local>

Hi
I'm trying to get a core dump out of nginx but it isn't working...

I have done the following:

Made directory with permissions:
drwxr-xr-x  2 root    root    4.0K 2011-05-26 00:32 nginx-core

changed config:
worker_rlimit_core  512M;
working_directory  /home/dcstaff/nginx-core/;

but no dump is made when it crashes:
2011/05/26 06:46:50 [alert] 20526#0: worker process 20547 exited on signal 11

Is there any other trick to it?





From Richard.Kearsley at m247.com  Thu May 26 12:11:09 2011
From: Richard.Kearsley at m247.com (Richard Kearsley)
Date: Thu, 26 May 2011 08:11:09 +0000
Subject: core dumps
In-Reply-To: <EFC5A19DBF3A2E489904733B467D990B131183@office-server2.m247.local>
References: <EFC5A19DBF3A2E489904733B467D990B131183@office-server2.m247.local>
Message-ID: <EFC5A19DBF3A2E489904733B467D990B1311D8@office-server2.m247.local>

I should say this is on linux - ubuntu 10.10



-----Original Message-----
From: Richard Kearsley [mailto:Richard.Kearsley at m247.com]
Sent: 26 May 2011 09:07
To: 'nginx at nginx.org'
Subject: core dumps



Hi

I'm trying to get a core dump out of nginx but it isn't working...



I have done the following:



Made directory with permissions:

drwxr-xr-x  2 root    root    4.0K 2011-05-26 00:32 nginx-core



changed config:

worker_rlimit_core  512M;

working_directory  /home/dcstaff/nginx-core/;



but no dump is made when it crashes:

2011/05/26 06:46:50 [alert] 20526#0: worker process 20547 exited on signal 11



Is there any other trick to it?









_______________________________________________

nginx mailing list

nginx at nginx.org

http://nginx.org/mailman/listinfo/nginx
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110526/9f20b6a6/attachment.html>

From mdounin at mdounin.ru  Thu May 26 12:26:08 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 26 May 2011 12:26:08 +0400
Subject: core dumps
In-Reply-To: <EFC5A19DBF3A2E489904733B467D990B131183@office-server2.m247.local>
References: <EFC5A19DBF3A2E489904733B467D990B131183@office-server2.m247.local>
Message-ID: <20110526082608.GE42265@mdounin.ru>

Hello!

On Thu, May 26, 2011 at 08:06:37AM +0000, Richard Kearsley wrote:

> Hi
> I'm trying to get a core dump out of nginx but it isn't working...
> 
> I have done the following:
> 
> Made directory with permissions:
> drwxr-xr-x  2 root    root    4.0K 2011-05-26 00:32 nginx-core
> 
> changed config:
> worker_rlimit_core  512M;
> working_directory  /home/dcstaff/nginx-core/;
> 
> but no dump is made when it crashes:
> 2011/05/26 06:46:50 [alert] 20526#0: worker process 20547 exited on signal 11
> 
> Is there any other trick to it?

Worker processes are usually run under non-root, so it's unlikely 
that write permissions for root will be enough.

Maxim Dounin


From nginx-forum at nginx.us  Thu May 26 12:34:51 2011
From: nginx-forum at nginx.us (chris.percol)
Date: Thu, 26 May 2011 04:34:51 -0400
Subject: redirect subdomain to internal location
In-Reply-To: <8762oyli45.wl%appa@perusio.net>
References: <8762oyli45.wl%appa@perusio.net>
Message-ID: <8539335ce68bd599dd41fd22d7ebae74.NginxMailingListEnglish@forum.nginx.org>

Ant?nio P. P. Almeida Wrote:
-------------------------------------------------------
> On 25 Mai 2011 19h14 WEST, nginx-forum at nginx.us
> wrote:
> 
> 
> >> location ~ /(?<clientkey>[^/]*)/ {
> >
> > Appa,
> >
> > Thanks again!
> >
> > My issue is that I don't know the value of
> clientkey, it could be
> > several hundred possibilities.
> >
> > Is there a way to extract the unknown clientkey
> from
> > /clientkey/some/more.htm?
> >
> > Many thanks,
> 
> That's completely generic. It's a named group that
> captures any char
> excepting '/'.
> 
> This implicitly creates a variable $clientkey that
> you can use.
> 
> Try it.
> --- appa
> 

Appa,

Thanks, no joy, but get it now. Think problem may be down to my perl
version, will try with ?P<clientname>.

Chris


> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200694,201071#msg-201071



From Richard.Kearsley at m247.com  Thu May 26 13:02:26 2011
From: Richard.Kearsley at m247.com (Richard Kearsley)
Date: Thu, 26 May 2011 09:02:26 +0000
Subject: core dumps
In-Reply-To: <20110526082608.GE42265@mdounin.ru>
References: <EFC5A19DBF3A2E489904733B467D990B131183@office-server2.m247.local>
	<20110526082608.GE42265@mdounin.ru>
Message-ID: <EFC5A19DBF3A2E489904733B467D990B131312@office-server2.m247.local>

Sorry, I chose a bad example
I have 3 machines where it's not dumping
The other 2 had the correct permissions already
Should it core dump for signal 11 of a worker?

> Worker processes are usually run under non-root, so it's unlikely 
> that write permissions for root will be enough.

_______________________________________________
nginx mailing list
nginx at nginx.org
http://nginx.org/mailman/listinfo/nginx


From nginx-forum at nginx.us  Thu May 26 14:11:10 2011
From: nginx-forum at nginx.us (slafs)
Date: Thu, 26 May 2011 06:11:10 -0400
Subject: IP as a server name
Message-ID: <d65f002d6daeef15d7feb3f2c8f595f0.NginxMailingListEnglish@forum.nginx.org>

Hi there!

I have a question about nginx configuration file. How can I configure a
server if I want my site to be available under a domain name and IP
address. I would like to have something like: 
http(s)://myhost.mydomain/foo/
and
http(s)://XXX.XXX.XXX.XXX/foo/
to be available simultaneously.

I noticed that it basically works but problem occurs on redirections
from my webapp (Nginx is a reverse proxy for it). It seems that on
redirects nginx is obtaining a server_name from OS configuration.

Can You please suggest some configuration?

I am using nginx 0.6.39 but will try to upgrade if it is necessary.

Regards

S?awek

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201092,201092#msg-201092



From josh at joshparker.us  Thu May 26 14:22:39 2011
From: josh at joshparker.us (=?utf-8?Q?Parker=2C_Joshua?=)
Date: Thu, 26 May 2011 06:22:39 -0400
Subject: Error with http_perl_module
In-Reply-To: <20110526074048.GI86377@sysoev.ru>
References: <7104EC025B8F4C558C84166030307E0C@joshparker.us>
	<20110526074048.GI86377@sysoev.ru>
Message-ID: <93C96B6191494951985C5928158923EA@joshparker.us>

Please see below.

Note (probably harmless): No library found for -lperl
-Wl,-E -fstack-protector -L/usr/local/lib -L/usr/lib/perl/5.10/CORE -ldl -lm -lpthread -lc -lcrypt

-- 
Parker, Joshua
Sent with Sparrow (http://www.sparrowmailapp.com/?sig)

On Thursday, May 26, 2011 at 3:40 AM, Igor Sysoev wrote:

> On Wed, May 25, 2011 at 11:03:52PM -0400, Parker, Joshua wrote:
> > I am trying to upgrade to 1.0.2, but I am getting an install error. Any help is greatly appreciated.
> 
> > -lcrypt -lpcre -lssl -lcrypto -ldl -lz \
> > -Wl,-E -fstack-protector -L/usr/local/lib -L/usr/lib/perl/5.10/CORE -ldl -lm -lpthread -lc -lcrypt
> > objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_init_interpreter':
> > /usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:516: undefined reference to `Perl_sys_init'
> 
> What does "perl -MExtUtils::Embed -e ldopts" show ?
> 
> 
> -- 
> Igor Sysoev
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org (mailto:nginx at nginx.org)
> http://nginx.org/mailman/listinfo/nginx

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110526/dfc8c18c/attachment.html>

From nginx-forum at nginx.us  Thu May 26 14:30:13 2011
From: nginx-forum at nginx.us (slafs)
Date: Thu, 26 May 2011 06:30:13 -0400
Subject: IP as a server name
In-Reply-To: <d65f002d6daeef15d7feb3f2c8f595f0.NginxMailingListEnglish@forum.nginx.org>
References: <d65f002d6daeef15d7feb3f2c8f595f0.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <e278fd3c476b9a2715ba299b88d8d9f8.NginxMailingListEnglish@forum.nginx.org>

Ok found it. Sorry.

server_name_in_redirect off;

did the trick.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201092,201100#msg-201100



From igor at sysoev.ru  Thu May 26 14:33:08 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Thu, 26 May 2011 14:33:08 +0400
Subject: Error with http_perl_module
In-Reply-To: <93C96B6191494951985C5928158923EA@joshparker.us>
References: <7104EC025B8F4C558C84166030307E0C@joshparker.us>
	<20110526074048.GI86377@sysoev.ru>
	<93C96B6191494951985C5928158923EA@joshparker.us>
Message-ID: <20110526103308.GA12142@sysoev.ru>

On Thu, May 26, 2011 at 06:22:39AM -0400, Parker, Joshua wrote:
> Please see below.
> 
> Note (probably harmless): No library found for -lperl
> -Wl,-E -fstack-protector -L/usr/local/lib -L/usr/lib/perl/5.10/CORE -ldl -lm -lpthread -lc -lcrypt

LOL, "probably harmless". libperl is the very library required for linking.
It's seems you have broken perl installation.

> -- 
> Parker, Joshua
> Sent with Sparrow (http://www.sparrowmailapp.com/?sig)
> 
> On Thursday, May 26, 2011 at 3:40 AM, Igor Sysoev wrote:
> 
> > On Wed, May 25, 2011 at 11:03:52PM -0400, Parker, Joshua wrote:
> > > I am trying to upgrade to 1.0.2, but I am getting an install error. Any help is greatly appreciated.
> > 
> > > -lcrypt -lpcre -lssl -lcrypto -ldl -lz \
> > > -Wl,-E -fstack-protector -L/usr/local/lib -L/usr/lib/perl/5.10/CORE -ldl -lm -lpthread -lc -lcrypt
> > > objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_init_interpreter':
> > > /usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:516: undefined reference to `Perl_sys_init'
> > 
> > What does "perl -MExtUtils::Embed -e ldopts" show ?


-- 
Igor Sysoev


From josh at joshparker.us  Thu May 26 14:36:08 2011
From: josh at joshparker.us (=?utf-8?Q?Parker=2C_Joshua?=)
Date: Thu, 26 May 2011 06:36:08 -0400
Subject: Error with http_perl_module
In-Reply-To: <20110526103308.GA12142@sysoev.ru>
References: <7104EC025B8F4C558C84166030307E0C@joshparker.us>
	<20110526074048.GI86377@sysoev.ru>
	<93C96B6191494951985C5928158923EA@joshparker.us>
	<20110526103308.GA12142@sysoev.ru>
Message-ID: <044BD18713004A6EA2A40D3106298F98@joshparker.us>

I don't remember installing perl, so the message may be a misdirection. How would I install perl on Ubuntu Lucid? I am not sure what perl package I should be looking for.



-- 
Parker, Joshua
Sent with Sparrow (http://www.sparrowmailapp.com/?sig)

On Thursday, May 26, 2011 at 6:33 AM, Igor Sysoev wrote:

> On Thu, May 26, 2011 at 06:22:39AM -0400, Parker, Joshua wrote:
> > Please see below.
> > 
> > Note (probably harmless): No library found for -lperl
> > -Wl,-E -fstack-protector -L/usr/local/lib -L/usr/lib/perl/5.10/CORE -ldl -lm -lpthread -lc -lcrypt
> 
> LOL, "probably harmless". libperl is the very library required for linking.
> It's seems you have broken perl installation.
> 
> > -- 
> > Parker, Joshua
> > Sent with Sparrow (http://www.sparrowmailapp.com/?sig)
> > 
> > On Thursday, May 26, 2011 at 3:40 AM, Igor Sysoev wrote:
> > 
> > > On Wed, May 25, 2011 at 11:03:52PM -0400, Parker, Joshua wrote:
> > > > I am trying to upgrade to 1.0.2, but I am getting an install error. Any help is greatly appreciated.
> > > 
> > > > -lcrypt -lpcre -lssl -lcrypto -ldl -lz \
> > > > -Wl,-E -fstack-protector -L/usr/local/lib -L/usr/lib/perl/5.10/CORE -ldl -lm -lpthread -lc -lcrypt
> > > > objs/src/http/modules/perl/ngx_http_perl_module.o: In function `ngx_http_perl_init_interpreter':
> > > > /usr/local/src/nginx-1.0.2/src/http/modules/perl/ngx_http_perl_module.c:516: undefined reference to `Perl_sys_init'
> > > 
> > > What does "perl -MExtUtils::Embed -e ldopts" show ?
> 
> 
> -- 
> Igor Sysoev
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org (mailto:nginx at nginx.org)
> http://nginx.org/mailman/listinfo/nginx

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110526/f2e6ab85/attachment.html>

From igor at sysoev.ru  Thu May 26 14:37:22 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Thu, 26 May 2011 14:37:22 +0400
Subject: core dumps
In-Reply-To: <EFC5A19DBF3A2E489904733B467D990B131312@office-server2.m247.local>
References: <EFC5A19DBF3A2E489904733B467D990B131183@office-server2.m247.local>
	<20110526082608.GE42265@mdounin.ru>
	<EFC5A19DBF3A2E489904733B467D990B131312@office-server2.m247.local>
Message-ID: <20110526103722.GB12142@sysoev.ru>

On Thu, May 26, 2011 at 09:02:26AM +0000, Richard Kearsley wrote:
> Sorry, I chose a bad example
> I have 3 machines where it's not dumping
> The other 2 had the correct permissions already
> Should it core dump for signal 11 of a worker?

Probably you need to set also:
echo 1 > /proc/sys/fs/suid_dumpable

> > Worker processes are usually run under non-root, so it's unlikely 
> > that write permissions for root will be enough.


-- 
Igor Sysoev


From igor at sysoev.ru  Thu May 26 14:39:47 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Thu, 26 May 2011 14:39:47 +0400
Subject: Error with http_perl_module
In-Reply-To: <044BD18713004A6EA2A40D3106298F98@joshparker.us>
References: <7104EC025B8F4C558C84166030307E0C@joshparker.us>
	<20110526074048.GI86377@sysoev.ru>
	<93C96B6191494951985C5928158923EA@joshparker.us>
	<20110526103308.GA12142@sysoev.ru>
	<044BD18713004A6EA2A40D3106298F98@joshparker.us>
Message-ID: <20110526103947.GC12142@sysoev.ru>

On Thu, May 26, 2011 at 06:36:08AM -0400, Parker, Joshua wrote:
> I don't remember installing perl, so the message may be a misdirection. How would I install perl on Ubuntu Lucid? I am not sure what perl package I should be looking for.

Last time I tried this on Debian, libper-dev package was required.


-- 
Igor Sysoev


From josh at joshparker.us  Thu May 26 14:40:17 2011
From: josh at joshparker.us (=?utf-8?Q?Parker=2C_Joshua?=)
Date: Thu, 26 May 2011 06:40:17 -0400
Subject: Error with http_perl_module
In-Reply-To: <20110526103947.GC12142@sysoev.ru>
References: <7104EC025B8F4C558C84166030307E0C@joshparker.us>
	<20110526074048.GI86377@sysoev.ru>
	<93C96B6191494951985C5928158923EA@joshparker.us>
	<20110526103308.GA12142@sysoev.ru>
	<044BD18713004A6EA2A40D3106298F98@joshparker.us>
	<20110526103947.GC12142@sysoev.ru>
Message-ID: <19866AB3B49849D686422D9BAD0E2071@joshparker.us>

Ok, thanks.



-- 
Parker, Joshua
Sent with Sparrow (http://www.sparrowmailapp.com/?sig)

On Thursday, May 26, 2011 at 6:39 AM, Igor Sysoev wrote:

> On Thu, May 26, 2011 at 06:36:08AM -0400, Parker, Joshua wrote:
> > I don't remember installing perl, so the message may be a misdirection. How would I install perl on Ubuntu Lucid? I am not sure what perl package I should be looking for.
> 
> Last time I tried this on Debian, libper-dev package was required.
> 
> 
> -- 
> Igor Sysoev
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org (mailto:nginx at nginx.org)
> http://nginx.org/mailman/listinfo/nginx

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110526/35bf2b13/attachment.html>

From nginx-forum at nginx.us  Thu May 26 14:47:38 2011
From: nginx-forum at nginx.us (chris.percol)
Date: Thu, 26 May 2011 06:47:38 -0400
Subject: redirect subdomain to internal location
In-Reply-To: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>
References: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <752f7a108532c2eed1e980fdc1806b01.NginxMailingListEnglish@forum.nginx.org>

Nearly there!

Thanks to Appa I have got this far...

location ~ /(?<clientkey>[^/]*)/ {

What I didn't consider was that sometimes the result I am testing for
may not have a proceeding '/'.

How I update the rexex to allow for /mytest as well as /mytest?

Many thanks,

Chris

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200694,201110#msg-201110



From appa at perusio.net  Thu May 26 15:04:09 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 26 May 2011 12:04:09 +0100
Subject: redirect subdomain to internal location
In-Reply-To: <752f7a108532c2eed1e980fdc1806b01.NginxMailingListEnglish@forum.nginx.org>
References: <98b315ad9cc32b0957db3ff6074d4799.NginxMailingListEnglish@forum.nginx.org>
	<752f7a108532c2eed1e980fdc1806b01.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <874o4hlo46.wl%appa@perusio.net>

On 26 Mai 2011 11h47 WEST, nginx-forum at nginx.us wrote:

> Nearly there!
>
> Thanks to Appa I have got this far...
>
> location ~ /(?<clientkey>[^/]*)/ {
>
> What I didn't consider was that sometimes the result I am testing
> for may not have a proceeding '/'.

location ~ ^/(?<clientkey>[^/]*) {
  (...)
}

All requests of the form domain.com/whatever are now matched. You're
capturing only the 1st component of the URI 'whatever'.

--- appa

 



From nginx-forum at nginx.us  Thu May 26 16:46:14 2011
From: nginx-forum at nginx.us (chris.percol)
Date: Thu, 26 May 2011 08:46:14 -0400
Subject: redirect subdomain to internal location
In-Reply-To: <874o4hlo46.wl%appa@perusio.net>
References: <874o4hlo46.wl%appa@perusio.net>
Message-ID: <23e4e44d2a87c6565eb231f9f0185a4b.NginxMailingListEnglish@forum.nginx.org>

Appa,

Thanks for your help and patience!

Chris

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,200694,201148#msg-201148



From nginx-forum at nginx.us  Thu May 26 17:46:47 2011
From: nginx-forum at nginx.us (the_traveller)
Date: Thu, 26 May 2011 09:46:47 -0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <c0c6df7135604bcee44a098e5b896a30@ruby-forum.com>
References: <c0c6df7135604bcee44a098e5b896a30@ruby-forum.com>
Message-ID: <978224f0ac741f3163cc698c0baa9103.NginxMailingListEnglish@forum.nginx.org>

the public_html dir is in
/usr/local/www

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,23868,201159#msg-201159



From igor at sysoev.ru  Thu May 26 17:51:42 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Thu, 26 May 2011 17:51:42 +0400
Subject: Rewrite rule for nginx from Apache
In-Reply-To: <978224f0ac741f3163cc698c0baa9103.NginxMailingListEnglish@forum.nginx.org>
References: <c0c6df7135604bcee44a098e5b896a30@ruby-forum.com>
	<978224f0ac741f3163cc698c0baa9103.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110526135142.GA37526@sysoev.ru>

On Thu, May 26, 2011 at 09:46:47AM -0400, the_traveller wrote:
> the public_html dir is in
> /usr/local/www

location /Joke/ {
    fastcgi_pass  backend;
    fastcgi_param SCRIPT_FILENAME /usr/local/www/public_html/jokes.php;
    fastcgi_param QUERY_STRING $args;
    inlcude       fastcgi_params;
}


-- 
Igor Sysoev


From lists at ruby-forum.com  Thu May 26 19:59:38 2011
From: lists at ruby-forum.com (John Travolota)
Date: Thu, 26 May 2011 17:59:38 +0200
Subject: kernel: nginx[18233]: segfault at 0000000000000008 rip
	000000000043edf8 rsp 00007fff34a21fa0 err
In-Reply-To: <d6ba00b08aeeb18ecc9a0873cdd53e5e@ruby-forum.com>
References: <d6ba00b08aeeb18ecc9a0873cdd53e5e@ruby-forum.com>
Message-ID: <4919c744e69acf0b9a463d19443be22f@ruby-forum.com>

It's a bit different segfault now: segfault at 0000000000000008 rip 
000000000043a5ab rsp 00007fff43146db0 error 4
Probably because i upgraded to version 1.3.I didn't do debugging,but i 
found out what exactly module causing segfault-it's geoip module.I 
noticed by reverting to ver seven first,and segfault occurred 
again,including when i changed conf to configure without SSI and 
SSL.Before this segfault error i didn't used geoip module.Perhaps is 
nginx overloaded due to a lot of request from redirected countries.Any 
idea how to fix it?

-- 
Posted via http://www.ruby-forum.com/.


From mdounin at mdounin.ru  Thu May 26 20:09:42 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 26 May 2011 20:09:42 +0400
Subject: kernel: nginx[18233]: segfault at 0000000000000008 rip
	000000000043edf8 rsp 00007fff34a21fa0 err
In-Reply-To: <4919c744e69acf0b9a463d19443be22f@ruby-forum.com>
References: <d6ba00b08aeeb18ecc9a0873cdd53e5e@ruby-forum.com>
	<4919c744e69acf0b9a463d19443be22f@ruby-forum.com>
Message-ID: <20110526160942.GF42265@mdounin.ru>

Hello!

On Thu, May 26, 2011 at 05:59:38PM +0200, John Travolota wrote:

> It's a bit different segfault now: segfault at 0000000000000008 rip 
> 000000000043a5ab rsp 00007fff43146db0 error 4
> Probably because i upgraded to version 1.3.I didn't do debugging,but i 
> found out what exactly module causing segfault-it's geoip module.I 
> noticed by reverting to ver seven first,and segfault occurred 
> again,including when i changed conf to configure without SSI and 
> SSL.Before this segfault error i didn't used geoip module.Perhaps is 
> nginx overloaded due to a lot of request from redirected countries.Any 
> idea how to fix it?

You may want to debug it further (see http://wiki.nginx.org/Debugging), 
but please note that MaxMind's GeoIP library (the one which geoip 
module uses) segfaults on incorrect/corrupted database[1].  
Checking if your MaxMind's database isn't corrupted is a good idea.

[1] And this is one of the reasons I recommend using geo module 
instead and always wonder if people use geoip.

Maxim Dounin


From nginx-forum at nginx.us  Thu May 26 23:22:36 2011
From: nginx-forum at nginx.us (benseb)
Date: Thu, 26 May 2011 15:22:36 -0400
Subject: Sanity check of my config - is it secure?
In-Reply-To: <3ce420be0701a81ea969389d395b6a49.NginxMailingListEnglish@forum.nginx.org>
References: <3ce420be0701a81ea969389d395b6a49.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <0df75b6b4e445800bb6bf833adb91008.NginxMailingListEnglish@forum.nginx.org>

Can anyone help with the above request, regarding checking if a filename
DOESNT match the whitelist above (block all other filetypes)

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199902,201268#msg-201268



From appa at perusio.net  Thu May 26 23:32:10 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 26 May 2011 20:32:10 +0100
Subject: Sanity check of my config - is it secure?
In-Reply-To: <0df75b6b4e445800bb6bf833adb91008.NginxMailingListEnglish@forum.nginx.org>
References: <3ce420be0701a81ea969389d395b6a49.NginxMailingListEnglish@forum.nginx.org>
	<0df75b6b4e445800bb6bf833adb91008.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87pqn5jm11.wl%appa@perusio.net>

On 26 Mai 2011 20h22 WEST, nginx-forum at nginx.us wrote:

Your're letting the reverse logical style of mod_rewrite and .htaccess
color your perception.

In Nginx things operate in a forward logical way:

1. Define which extensions you want to allow, e.g.:

location ~* \.(?:jpe?g|png|ico|gif|css|js|) {
  # serve the files
}

location ~* (which extensions are going to be blocked) {
 return 444;
}

Mind you that relying solely on the file extension is a rather weak
way of filtering files. You can tamper the file magic number quite
easily.

> Can anyone help with the above request, regarding checking if a
> filename DOESNT match the whitelist above (block all other
> filetypes)

--- appa



From nginx-forum at nginx.us  Thu May 26 23:36:51 2011
From: nginx-forum at nginx.us (benseb)
Date: Thu, 26 May 2011 15:36:51 -0400
Subject: Sanity check of my config - is it secure?
In-Reply-To: <87pqn5jm11.wl%appa@perusio.net>
References: <87pqn5jm11.wl%appa@perusio.net>
Message-ID: <984ac956e570b8e9b5e5b22ee356cd9b.NginxMailingListEnglish@forum.nginx.org>

Thanks, in your opinion what's the best way to approach this? I
basically want to ensure that our static.domain.com subdomain ONLY
servers image/js/css files.

Whilst I have set 'location' for only the folders which have images, etc
in, I want to ensure that if someone put a script into one of those
directories, it would not be executed.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199902,201276#msg-201276



From appa at perusio.net  Thu May 26 23:46:50 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 26 May 2011 20:46:50 +0100
Subject: Sanity check of my config - is it secure?
In-Reply-To: <984ac956e570b8e9b5e5b22ee356cd9b.NginxMailingListEnglish@forum.nginx.org>
References: <87pqn5jm11.wl%appa@perusio.net>
	<984ac956e570b8e9b5e5b22ee356cd9b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87oc2pjlcl.wl%appa@perusio.net>

On 26 Mai 2011 20h36 WEST, nginx-forum at nginx.us wrote:

> Thanks, in your opinion what's the best way to approach this? I
> basically want to ensure that our static.domain.com subdomain ONLY
> servers image/js/css files.
>
> Whilst I have set 'location' for only the folders which have images,
> etc in, I want to ensure that if someone put a script into one of
> those directories, it would not be executed.

If I understood correctly: there's no need to worry then. If there's
no embedded interpreter (Perl/Lua) or a fastcgi backend configured on
that vhost. There's no way that someone will be able to run a script
from the web there.

--- appa



From nginx-forum at nginx.us  Thu May 26 23:49:41 2011
From: nginx-forum at nginx.us (benseb)
Date: Thu, 26 May 2011 15:49:41 -0400
Subject: Sanity check of my config - is it secure?
In-Reply-To: <87oc2pjlcl.wl%appa@perusio.net>
References: <87oc2pjlcl.wl%appa@perusio.net>
Message-ID: <4eeedadeefe2d50b4c7c7b3e3f487255.NginxMailingListEnglish@forum.nginx.org>

I presume in that case, if there was a script and no interpreter, it
would just display the contents of the file - which again could be a
security risk.

So I need to perhaps limit the mime types that can be served, or the
file extensions - which ever is most secure?

I basically don't want a php script in that directory being served and
the source code being visible

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199902,201282#msg-201282



From appa at perusio.net  Thu May 26 23:59:31 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 26 May 2011 20:59:31 +0100
Subject: Sanity check of my config - is it secure?
In-Reply-To: <4eeedadeefe2d50b4c7c7b3e3f487255.NginxMailingListEnglish@forum.nginx.org>
References: <87oc2pjlcl.wl%appa@perusio.net>
	<4eeedadeefe2d50b4c7c7b3e3f487255.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87mxi9jkrg.wl%appa@perusio.net>

On 26 Mai 2011 20h49 WEST, nginx-forum at nginx.us wrote:


> I presume in that case, if there was a script and no interpreter, it
> would just display the contents of the file - which again could be a
> security risk.

> So I need to perhaps limit the mime types that can be served, or the
> file extensions - which ever is most secure?


Do both. That's my advice. Regarding the later, something along the
lines of:


location ^~ /static_files_dir/ {
  
  location ~* /static_files_dir/.+\.(?:jpe?g|png|gif|ico|css|js)$ {
     expires 30d;
  }

  location ~* /static_files_dir/.*\.php$ {
     return 444; # return an empty response for a php file
  }  
}

--- appa



From nunomagalhaes at eu.ipp.pt  Fri May 27 00:04:31 2011
From: nunomagalhaes at eu.ipp.pt (=?UTF-8?Q?Nuno_Magalh=C3=A3es?=)
Date: Thu, 26 May 2011 21:04:31 +0100
Subject: Sanity check of my config - is it secure?
In-Reply-To: <87mxi9jkrg.wl%appa@perusio.net>
References: <87oc2pjlcl.wl%appa@perusio.net>
	<4eeedadeefe2d50b4c7c7b3e3f487255.NginxMailingListEnglish@forum.nginx.org>
	<87mxi9jkrg.wl%appa@perusio.net>
Message-ID: <BANLkTi=QBEocJR+kESarLOuJP1SBpLZPew@mail.gmail.com>

Hi

On Thu, May 26, 2011 at 20:59, Ant?nio P. P. Almeida <appa at perusio.net> wrote:
> return 444; # return an empty response for a php file

AFAIK 444 just closes the connection, sending nothing.
http://wiki.nginx.org/HttpRewriteModule#return

Or does it actually send an empty response (without headers)?

-- 
Mars 2 Stay!
http://xkcd.com/801/
/etc


From nginx-forum at nginx.us  Fri May 27 00:10:23 2011
From: nginx-forum at nginx.us (benseb)
Date: Thu, 26 May 2011 16:10:23 -0400
Subject: Sanity check of my config - is it secure?
In-Reply-To: <4eeedadeefe2d50b4c7c7b3e3f487255.NginxMailingListEnglish@forum.nginx.org>
References: <87oc2pjlcl.wl%appa@perusio.net>
	<4eeedadeefe2d50b4c7c7b3e3f487255.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <3538d29804bc922548e34af913cb5963.NginxMailingListEnglish@forum.nginx.org>

Thanks people

So there's no way to say "If the file ISNT a jpeg/gif/css/js" deny. The
only way is to say 'if .php' deny, 'if .txt deny' etc?

I'd prefer to whitelist the files i DO want to return and block
everything else, incase I forget something to block?

Ben

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199902,201289#msg-201289



From appa at perusio.net  Fri May 27 00:18:31 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 26 May 2011 21:18:31 +0100
Subject: Sanity check of my config - is it secure?
In-Reply-To: <3538d29804bc922548e34af913cb5963.NginxMailingListEnglish@forum.nginx.org>
References: <87oc2pjlcl.wl%appa@perusio.net>
	<4eeedadeefe2d50b4c7c7b3e3f487255.NginxMailingListEnglish@forum.nginx.org>
	<3538d29804bc922548e34af913cb5963.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87lixtjjvs.wl%appa@perusio.net>

On 26 Mai 2011 21h10 WEST, nginx-forum at nginx.us wrote:

> Thanks people
>
> So there's no way to say "If the file ISNT a jpeg/gif/css/js"
> deny. The only way is to say 'if .php' deny, 'if .txt deny' etc?
>
> I'd prefer to whitelist the files i DO want to return and block
> everything else, incase I forget something to block?
>

Try this: 

location ~* /static_files_dir/(?[^.]*)\.(?<extension>.*)$ {
  if ($extension !~ (?:jpe?g|png|gif|ico|css|js)) {
    return 444;
  }
}

It's rather ugly :(

--- appa



From appa at perusio.net  Fri May 27 00:20:57 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 26 May 2011 21:20:57 +0100
Subject: Sanity check of my config - is it secure?
In-Reply-To: <3538d29804bc922548e34af913cb5963.NginxMailingListEnglish@forum.nginx.org>
References: <87oc2pjlcl.wl%appa@perusio.net>
	<4eeedadeefe2d50b4c7c7b3e3f487255.NginxMailingListEnglish@forum.nginx.org>
	<3538d29804bc922548e34af913cb5963.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87k4ddjjrq.wl%appa@perusio.net>

On 26 Mai 2011 21h10 WEST, nginx-forum at nginx.us wrote:

> Thanks people
>
> So there's no way to say "If the file ISNT a jpeg/gif/css/js"
> deny. The only way is to say 'if .php' deny, 'if .txt deny' etc?
>
> I'd prefer to whitelist the files i DO want to return and block
> everything else, incase I forget something to block?
>

Try this: 

location ~* /static_files_dir/(?:[^.]*)\.(?<extension>.*)$ {
  if ($extension !~ (jpe?g|png|gif|ico|css|js)) {
    return 444;
  }
}

It's rather ugly :(

--- appa

PS: No need to use ?: in the if condition.


From appa at perusio.net  Fri May 27 00:23:33 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 26 May 2011 21:23:33 +0100
Subject: Sanity check of my config - is it secure?
In-Reply-To: <3538d29804bc922548e34af913cb5963.NginxMailingListEnglish@forum.nginx.org>
References: <87oc2pjlcl.wl%appa@perusio.net>
	<4eeedadeefe2d50b4c7c7b3e3f487255.NginxMailingListEnglish@forum.nginx.org>
	<3538d29804bc922548e34af913cb5963.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87ipsxjjne.wl%appa@perusio.net>

On 26 Mai 2011 21h10 WEST, nginx-forum at nginx.us wrote:

> Thanks people
>
> So there's no way to say "If the file ISNT a jpeg/gif/css/js"
> deny. The only way is to say 'if .php' deny, 'if .txt deny' etc?
>
> I'd prefer to whitelist the files i DO want to return and block
> everything else, incase I forget something to block?
>

Try this: 

location ~* /static_files_dir/(?:[^.]*)\.(?<extension>.*)$ {
  if ($extension !~ (jpe?g|png|gif|ico|css|js)) {
    return 444;
  }
}

It's rather ugly :(

--- appa

PS: No need to use ?: in the if condition.

PPS: Note that if the filenames have dots '.' in them the above regex
will fail to capture the extension correctly. You're opening a can of
worms config wise when going down this path.


From nginx-forum at nginx.us  Fri May 27 00:30:53 2011
From: nginx-forum at nginx.us (benseb)
Date: Thu, 26 May 2011 16:30:53 -0400
Subject: Sanity check of my config - is it secure?
In-Reply-To: <87k4ddjjrq.wl%appa@perusio.net>
References: <87k4ddjjrq.wl%appa@perusio.net>
Message-ID: <56b117011b76c1e7bf01d0db1635ba95.NginxMailingListEnglish@forum.nginx.org>

Thanks for the advice

Seems strange that this isn't an easy thing to do. After all, ALL
security advise always recommends whitelisting what you want and denying
everything else!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199902,201299#msg-201299



From appa at perusio.net  Fri May 27 00:42:55 2011
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 26 May 2011 21:42:55 +0100
Subject: Sanity check of my config - is it secure?
In-Reply-To: <56b117011b76c1e7bf01d0db1635ba95.NginxMailingListEnglish@forum.nginx.org>
References: <87k4ddjjrq.wl%appa@perusio.net>
	<56b117011b76c1e7bf01d0db1635ba95.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87hb8hjir4.wl%appa@perusio.net>

On 26 Mai 2011 21h30 WEST, nginx-forum at nginx.us wrote:

> Thanks for the advice
>
> Seems strange that this isn't an easy thing to do. After all, ALL
> security advise always recommends whitelisting what you want and
> denying everything else!

The config with two regex locations nested did that. But if you're
asking for a *catch all* regex that blocks every other extension
besides css, js, &c, then you're thinking in terms of the
complement of the set of allowed extensions.

It's easier to enunciate the negative than the positivem due to the
fact that you're "searching" a wide space.

--- appa




> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,199902,201299#msg-201299
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx


From astarr at wiredquote.com  Fri May 27 01:06:28 2011
From: astarr at wiredquote.com (Aaron Starr)
Date: Thu, 26 May 2011 14:06:28 -0700
Subject: Sanity check of my config - is it secure?
In-Reply-To: <87hb8hjir4.wl%appa@perusio.net>
References: <87k4ddjjrq.wl%appa@perusio.net>
	<56b117011b76c1e7bf01d0db1635ba95.NginxMailingListEnglish@forum.nginx.org>
	<87hb8hjir4.wl%appa@perusio.net>
Message-ID: <BANLkTimGON8KLioQyneatKYtu1JrLRPDHA@mail.gmail.com>

Without actually testing anything, can't you do something like this:


location ~* \.(jpe?g|png|gif)$ {
}

location / {
  return 444;
}

I.e., if the extension looks like an image, handle it normally. Otherwise,
in the normal case, return 444 (or whatever error code is appropriate).

Aaron


On Thu, May 26, 2011 at 1:42 PM, Ant?nio P. P. Almeida <appa at perusio.net>wrote:

> On 26 Mai 2011 21h30 WEST, nginx-forum at nginx.us wrote:
>
> > Thanks for the advice
> >
> > Seems strange that this isn't an easy thing to do. After all, ALL
> > security advise always recommends whitelisting what you want and
> > denying everything else!
>
> The config with two regex locations nested did that. But if you're
> asking for a *catch all* regex that blocks every other extension
> besides css, js, &c, then you're thinking in terms of the
> complement of the set of allowed extensions.
>
> It's easier to enunciate the negative than the positivem due to the
> fact that you're "searching" a wide space.
>
> --- appa
>
>
>
>
> > Posted at Nginx Forum:
> > http://forum.nginx.org/read.php?2,199902,201299#msg-201299
> >
> >
> > _______________________________________________
> > nginx mailing list
> > nginx at nginx.org
> > http://nginx.org/mailman/listinfo/nginx
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110526/30c3922d/attachment.html>

From nginx-forum at nginx.us  Fri May 27 01:09:04 2011
From: nginx-forum at nginx.us (benseb)
Date: Thu, 26 May 2011 17:09:04 -0400
Subject: Sanity check of my config - is it secure?
In-Reply-To: <BANLkTimGON8KLioQyneatKYtu1JrLRPDHA@mail.gmail.com>
References: <BANLkTimGON8KLioQyneatKYtu1JrLRPDHA@mail.gmail.com>
Message-ID: <f903b8b367cf4421d2052697033fd6ad.NginxMailingListEnglish@forum.nginx.org>

That would be a nice simple solution. I'll give that a go!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,199902,201314#msg-201314



From nginx-forum at nginx.us  Fri May 27 04:28:56 2011
From: nginx-forum at nginx.us (adamchal)
Date: Thu, 26 May 2011 20:28:56 -0400
Subject: cache gzipped content with proxy_pass
In-Reply-To: <abd47169bb79f08e7851998d6da1c5f6.NginxMailingListEnglish@forum.nginx.org>
References: <abd47169bb79f08e7851998d6da1c5f6.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <f6ccb96ccbd35ddb892367d10e102b5f.NginxMailingListEnglish@forum.nginx.org>

I've been following this discussion and I'm anxiously waiting for some
resolution.  I think Maxim's approach is right: implementing a
gzip_cache filter that could be used for caching any content.  This
would allow for much more flexibility than just using it for
proxy_pass/proxy_cache.

This could effectively eliminate the gzip_static filter.  If you have
plain HTML served and "gzip on;" plus something like "gzip_cache on;"
all of your plain text content would be served using the cached gzip
content.  Similarly, you would be able to use this for proxy_pass,
proxy_store, fastcgi_pass, etc.  It wouldn't matter since the caching of
the gzip'd content would be stored in the gzip_cache filter and
negotiated directly with the client.

There's a lot of chatter about using Varnish or Lighty or <insert
another web/cache server here> to give Nginx pre-gzipped data in a proxy
response.  This sounds absolutely horrible.  The bottom line is that if
Nginx is the webserver that is dealing with the client directly, it
should then be Nginx's responsibility to serve the appropriate version
(gzip'd or plain text) content based on the client's request headers. 
That being said, Nginx should be managing it's own cache of filtered
items.

The same goes for the image_filter.  Currently, I use the image_filter
for a project and I'm bummed out that Nginx is scaling the same image
over and over each time it is requested.  An image_filter_cache would be
another great feature.

Am I on the right track?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,120314,201370#msg-201370



From baalchina at gmail.com  Fri May 27 07:26:24 2011
From: baalchina at gmail.com (baalchina)
Date: Fri, 27 May 2011 11:26:24 +0800
Subject: nginx 1.0.3 with php-fpm, php report 404 error
Message-ID: <BANLkTinK35b14m2gsOOgKdPxpj7J_EzP4A@mail.gmail.com>

Hi everyone

I am using nginx 1.0.3 under CentOS 5.6, with php-fpm.

nginx works fine with static file, such as html,txt. But not with php file.

I wrote a phpinfo.php file in my webserver, my browser  and nginx log both
report an 404 error.

I changed nothing but only add below in my nginx.conf file:

*location ~ \.php$ {

fastcgi_pass   127.0.0.1:9000;
fastcgi_index  index.php;
include        /usr/local/nginx/conf/fastcgi.conf;
}*

this is fastcgi.conf:
*
fastcgi_param  SCRIPT_FILENAME    $document_root$fastcgi_script_name;


fastcgi_param  QUERY_STRING       $query_string;
fastcgi_param  REQUEST_METHOD     $request_method;
fastcgi_param  CONTENT_TYPE       $content_type;
fastcgi_param  CONTENT_LENGTH     $content_length;

fastcgi_param  SCRIPT_NAME        $fastcgi_script_name;
fastcgi_param  REQUEST_URI        $request_uri;
fastcgi_param  DOCUMENT_URI       $document_uri;
fastcgi_param  DOCUMENT_ROOT      $document_root;
fastcgi_param  SERVER_PROTOCOL    $server_protocol;

fastcgi_param  GATEWAY_INTERFACE  CGI/1.1;
fastcgi_param  SERVER_SOFTWARE    nginx/$nginx_version;

fastcgi_param  REMOTE_ADDR        $remote_addr;
fastcgi_param  REMOTE_PORT        $remote_port;
fastcgi_param  SERVER_ADDR        $server_addr;
fastcgi_param  SERVER_PORT        $server_port;
fastcgi_param  SERVER_NAME        $server_name;

# PHP only, required if PHP was built with --enable-force-cgi-redirect
#fastcgi_param  REDIRECT_STATUS    200;*

and php-fm works fine:
*[root at cacti database]# more /data/logs/php-fpm.log
[27-May-2011 10:04:51] NOTICE: configuration file
/usr/local/php/etc/php-fpm.conf test is successful

[27-May-2011 10:07:14] NOTICE: using inherited socket fd=6, "127.0.0.1:9000"
[27-May-2011 10:07:14] NOTICE: fpm is running, pid 25610
[27-May-2011 10:07:14] NOTICE: ready to handle connections*

and this is nginx access log:


*172.17.188.133 - - [27/May/2011:10:46:37 +0800] "GET /info.php HTTP/1.1"
404 5 "-" "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:2.0.1) Gecko/20100101
Firefox/4.0.
1"*

so, what is the problely problem of this 404 error?

thanks!
-- 
from:baalchina
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/0e6b224b/attachment.html>

From edhoprima at gmail.com  Fri May 27 07:43:04 2011
From: edhoprima at gmail.com (Edho P Arief)
Date: Fri, 27 May 2011 10:43:04 +0700
Subject: nginx 1.0.3 with php-fpm, php report 404 error
In-Reply-To: <BANLkTinK35b14m2gsOOgKdPxpj7J_EzP4A@mail.gmail.com>
References: <BANLkTinK35b14m2gsOOgKdPxpj7J_EzP4A@mail.gmail.com>
Message-ID: <BANLkTikgTqwOD8RJah-Gf27+oqbMoGjHfQ@mail.gmail.com>

On Fri, May 27, 2011 at 10:26 AM, baalchina <baalchina at gmail.com> wrote:
> Hi everyone
>
> I am using nginx 1.0.3 under CentOS 5.6, with php-fpm.
>
> nginx works fine with static file, such as html,txt. But not with php file.
>
> I wrote a phpinfo.php file in my webserver, my browser? and nginx log both
> report an 404 error.
>

did you set root parameter in location block (eg. location / { .. })
or in server block (server { ... })? You must set it in server block
if you want the location php to use the root or else it'll use default
root. Checking error log could also show you where nginx tried to find
the file.


From mat999 at gmail.com  Fri May 27 07:48:21 2011
From: mat999 at gmail.com (SplitIce)
Date: Fri, 27 May 2011 13:48:21 +1000
Subject: cache gzipped content with proxy_pass
In-Reply-To: <f6ccb96ccbd35ddb892367d10e102b5f.NginxMailingListEnglish@forum.nginx.org>
References: <abd47169bb79f08e7851998d6da1c5f6.NginxMailingListEnglish@forum.nginx.org>
	<f6ccb96ccbd35ddb892367d10e102b5f.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTi=Bihfw5cYZDS7M=B9nNs0OgGdkWA@mail.gmail.com>

You can cache the image filter using either lua or through a proxy pass to
yourself.

But yes I agree with your points.

On Fri, May 27, 2011 at 10:28 AM, adamchal <nginx-forum at nginx.us> wrote:

> I've been following this discussion and I'm anxiously waiting for some
> resolution.  I think Maxim's approach is right: implementing a
> gzip_cache filter that could be used for caching any content.  This
> would allow for much more flexibility than just using it for
> proxy_pass/proxy_cache.
>
> This could effectively eliminate the gzip_static filter.  If you have
> plain HTML served and "gzip on;" plus something like "gzip_cache on;"
> all of your plain text content would be served using the cached gzip
> content.  Similarly, you would be able to use this for proxy_pass,
> proxy_store, fastcgi_pass, etc.  It wouldn't matter since the caching of
> the gzip'd content would be stored in the gzip_cache filter and
> negotiated directly with the client.
>
> There's a lot of chatter about using Varnish or Lighty or <insert
> another web/cache server here> to give Nginx pre-gzipped data in a proxy
> response.  This sounds absolutely horrible.  The bottom line is that if
> Nginx is the webserver that is dealing with the client directly, it
> should then be Nginx's responsibility to serve the appropriate version
> (gzip'd or plain text) content based on the client's request headers.
> That being said, Nginx should be managing it's own cache of filtered
> items.
>
> The same goes for the image_filter.  Currently, I use the image_filter
> for a project and I'm bummed out that Nginx is scaling the same image
> over and over each time it is requested.  An image_filter_cache would be
> another great feature.
>
> Am I on the right track?
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,120314,201370#msg-201370
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>



-- 
Warez Scene <http://thewarezscene.org> Free Rapidshare
Downloads<http://www.nexusddl.com>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/3d97bfcf/attachment.html>

From julyclyde at gmail.com  Fri May 27 07:55:25 2011
From: julyclyde at gmail.com (=?GB2312?B?yM7P/sDa?=)
Date: Fri, 27 May 2011 11:55:25 +0800
Subject: nginx 1.0.3 with php-fpm, php report 404 error
In-Reply-To: <BANLkTinK35b14m2gsOOgKdPxpj7J_EzP4A@mail.gmail.com>
References: <BANLkTinK35b14m2gsOOgKdPxpj7J_EzP4A@mail.gmail.com>
Message-ID: <1ADBE276-9B97-47BE-892E-A8506035D1DA@gmail.com>

location ~ \.php$   should be a part of  location /
Or else, nginx cannot determine $document_root


? 2011-5-27???11:26? baalchina ???

> Hi everyone
> 
> I am using nginx 1.0.3 under CentOS 5.6, with php-fpm.
> 
> nginx works fine with static file, such as html,txt. But not with php file.
> 
> I wrote a phpinfo.php file in my webserver, my browser  and nginx log both report an 404 error.
> 
> I changed nothing but only add below in my nginx.conf file:
> 
> location ~ \.php$ {
> 
> fastcgi_pass   127.0.0.1:9000;
> fastcgi_index  index.php;
> include        /usr/local/nginx/conf/fastcgi.conf;
> }
> 
> this is fastcgi.conf:
> 
> fastcgi_param  SCRIPT_FILENAME    $document_root$fastcgi_script_name;
> 
> 
> fastcgi_param  QUERY_STRING       $query_string;
> fastcgi_param  REQUEST_METHOD     $request_method;
> fastcgi_param  CONTENT_TYPE       $content_type;
> fastcgi_param  CONTENT_LENGTH     $content_length;
> 
> fastcgi_param  SCRIPT_NAME        $fastcgi_script_name;
> fastcgi_param  REQUEST_URI        $request_uri;
> fastcgi_param  DOCUMENT_URI       $document_uri;
> fastcgi_param  DOCUMENT_ROOT      $document_root;
> fastcgi_param  SERVER_PROTOCOL    $server_protocol;
> 
> fastcgi_param  GATEWAY_INTERFACE  CGI/1.1;
> fastcgi_param  SERVER_SOFTWARE    nginx/$nginx_version;
> 
> fastcgi_param  REMOTE_ADDR        $remote_addr;
> fastcgi_param  REMOTE_PORT        $remote_port;
> fastcgi_param  SERVER_ADDR        $server_addr;
> fastcgi_param  SERVER_PORT        $server_port;
> fastcgi_param  SERVER_NAME        $server_name;
> 
> # PHP only, required if PHP was built with --enable-force-cgi-redirect
> #fastcgi_param  REDIRECT_STATUS    200;
> 
> and php-fm works fine:
> [root at cacti database]# more /data/logs/php-fpm.log 
> [27-May-2011 10:04:51] NOTICE: configuration file /usr/local/php/etc/php-fpm.conf test is successful
> 
> [27-May-2011 10:07:14] NOTICE: using inherited socket fd=6, "127.0.0.1:9000"
> [27-May-2011 10:07:14] NOTICE: fpm is running, pid 25610
> [27-May-2011 10:07:14] NOTICE: ready to handle connections
> 
> and this is nginx access log:
> 
> 
> 172.17.188.133 - - [27/May/2011:10:46:37 +0800] "GET /info.php HTTP/1.1" 404 5 "-" "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:2.0.1) Gecko/20100101 Firefox/4.0.
> 1"
> 
> so, what is the problely problem of this 404 error?
> 
> thanks!
> -- 
> from:baalchina
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/74673e1b/attachment.html>

From baalchina at gmail.com  Fri May 27 08:11:48 2011
From: baalchina at gmail.com (baalchina)
Date: Fri, 27 May 2011 12:11:48 +0800
Subject: nginx 1.0.3 with php-fpm, php report 404 error
In-Reply-To: <1ADBE276-9B97-47BE-892E-A8506035D1DA@gmail.com>
References: <BANLkTinK35b14m2gsOOgKdPxpj7J_EzP4A@mail.gmail.com>
	<1ADBE276-9B97-47BE-892E-A8506035D1DA@gmail.com>
Message-ID: <BANLkTi=3zyWb0dtioCymyBLf4nwxJGNsxA@mail.gmail.com>

thanks!

I moved location ~\.php to location \,works fine now.

2011/5/27 ??? <julyclyde at gmail.com>

> location ~ \.php$   should be a part of  location /
> Or else, nginx cannot determine $document_root
>
>
> ? 2011-5-27???11:26? baalchina ???
>
> Hi everyone
>
> I am using nginx 1.0.3 under CentOS 5.6, with php-fpm.
>
> nginx works fine with static file, such as html,txt. But not with php file.
>
> I wrote a phpinfo.php file in my webserver, my browser  and nginx log both
> report an 404 error.
>
> I changed nothing but only add below in my nginx.conf file:
>
> *location ~ \.php$ {
>
> fastcgi_pass   127.0.0.1:9000;
> fastcgi_index  index.php;
> include        /usr/local/nginx/conf/fastcgi.conf;
> }*
>
> this is fastcgi.conf:
> *
> fastcgi_param  SCRIPT_FILENAME    $document_root$fastcgi_script_name;
>
>
> fastcgi_param  QUERY_STRING       $query_string;
> fastcgi_param  REQUEST_METHOD     $request_method;
> fastcgi_param  CONTENT_TYPE       $content_type;
> fastcgi_param  CONTENT_LENGTH     $content_length;
>
> fastcgi_param  SCRIPT_NAME        $fastcgi_script_name;
> fastcgi_param  REQUEST_URI        $request_uri;
> fastcgi_param  DOCUMENT_URI       $document_uri;
> fastcgi_param  DOCUMENT_ROOT      $document_root;
> fastcgi_param  SERVER_PROTOCOL    $server_protocol;
>
> fastcgi_param  GATEWAY_INTERFACE  CGI/1.1;
> fastcgi_param  SERVER_SOFTWARE    nginx/$nginx_version;
>
> fastcgi_param  REMOTE_ADDR        $remote_addr;
> fastcgi_param  REMOTE_PORT        $remote_port;
> fastcgi_param  SERVER_ADDR        $server_addr;
> fastcgi_param  SERVER_PORT        $server_port;
> fastcgi_param  SERVER_NAME        $server_name;
>
> # PHP only, required if PHP was built with --enable-force-cgi-redirect
> #fastcgi_param  REDIRECT_STATUS    200;*
>
> and php-fm works fine:
> *[root at cacti database]# more /data/logs/php-fpm.log
> [27-May-2011 10:04:51] NOTICE: configuration file
> /usr/local/php/etc/php-fpm.conf test is successful
>
> [27-May-2011 10:07:14] NOTICE: using inherited socket fd=6, "
> 127.0.0.1:9000"
> [27-May-2011 10:07:14] NOTICE: fpm is running, pid 25610
> [27-May-2011 10:07:14] NOTICE: ready to handle connections*
>
> and this is nginx access log:
>
>
> *172.17.188.133 - - [27/May/2011:10:46:37 +0800] "GET /info.php HTTP/1.1"
> 404 5 "-" "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:2.0.1) Gecko/20100101
> Firefox/4.0.
> 1"*
>
> so, what is the problely problem of this 404 error?
>
> thanks!
> --
> from:baalchina
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>


-- 
from:baalchina
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/90367635/attachment.html>

From josh at joshparker.us  Fri May 27 08:17:10 2011
From: josh at joshparker.us (=?utf-8?Q?Parker=2C_Joshua?=)
Date: Fri, 27 May 2011 00:17:10 -0400
Subject: SMTP or POP3 Authentication
Message-ID: <6C6276B0EF904519A3133AB1D85ED644@joshparker.us>

I am trying to use smtp connection for an email marketing software and would like to use smtp or pop3, but I am not sure where to start. I have mailauth.pm in /root/auth/ but I am not sure if that is write. If the previous is correct, is this also correct in nginx.conf:

mail { auth_http 127.0.0.1:80/auth; pop3_capabilities "TOP" "USER";
 smtp_auth plain login; server { listen 110; protocol pop3; proxy on; } }

If none of the above is correct, can you point me in the write direction or documentation on how to accomplish this? Thanks.



-- 
Parker, Joshua
Sent with Sparrow (http://www.sparrowmailapp.com/?sig)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/0a8843a4/attachment.html>

From nginx-forum at nginx.us  Fri May 27 08:49:30 2011
From: nginx-forum at nginx.us (speedfirst)
Date: Fri, 27 May 2011 00:49:30 -0400
Subject: How to add cookie when proxy_pass via programming?
Message-ID: <3dc5bd76bbf522bacd924a5f91ecc262.NginxMailingListEnglish@forum.nginx.org>

Hey, I want to add custom cookie when proxy pass to the upstream in my
custom module. I want to implement like what this config does:

proxy_pass http://servers;
proxy_add_header Cookie "$http_cookie;KEY=VALUE";

But sinc the VALUE is dynamically generated so I need to add this cookie
via programming.

How to do that?

Thanks very much.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201413,201413#msg-201413



From igor at sysoev.ru  Fri May 27 09:39:10 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 09:39:10 +0400
Subject: Sanity check of my config - is it secure?
In-Reply-To: <BANLkTimOiuBPh7fYqMD+Cd1XKXskwJTAQA@mail.gmail.com>
References: <3ce420be0701a81ea969389d395b6a49.NginxMailingListEnglish@forum.nginx.org>
	<BANLkTimOiuBPh7fYqMD+Cd1XKXskwJTAQA@mail.gmail.com>
Message-ID: <20110527053910.GE79260@sysoev.ru>

On Sun, May 22, 2011 at 05:48:42PM +0100, Nuno Magalh?es wrote:

> > ? ? ? ? location ^~ /applets/{
> > ? ? ? ? ? ? ? root /home/vhosts/x.com/httpdocs;
> > ? ? ? ? ? ? ? expires max;
> > ? ? ? ?}
> > ? ? ? ?location ^~ /css/{
> > ? ? ? ? ? ? ? root /home/vhosts/x.com/httpdocs;
> > ? ? ? ? ? ? ? expires max;
> > ? ? ? ?}
> 
> And merge these locations (as well as others) perhaps? Something like
> ^~ /(applets|css)/

The lesser regex locaitons you have in configuraiton is the better.
The best case is zero regex locations.


-- 
Igor Sysoev


From igor at sysoev.ru  Fri May 27 09:41:02 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 09:41:02 +0400
Subject: Sanity check of my config - is it secure?
In-Reply-To: <3538d29804bc922548e34af913cb5963.NginxMailingListEnglish@forum.nginx.org>
References: <87oc2pjlcl.wl%appa@perusio.net>
	<4eeedadeefe2d50b4c7c7b3e3f487255.NginxMailingListEnglish@forum.nginx.org>
	<3538d29804bc922548e34af913cb5963.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110527054102.GF79260@sysoev.ru>

On Thu, May 26, 2011 at 04:10:23PM -0400, benseb wrote:
> Thanks people
> 
> So there's no way to say "If the file ISNT a jpeg/gif/css/js" deny. The
> only way is to say 'if .php' deny, 'if .txt deny' etc?
> 
> I'd prefer to whitelist the files i DO want to return and block
> everything else, incase I forget something to block?

As it was already suggested:

location ~* \.(jpe?g|png|gif)$ {
   root  ...
}

location / {
   return 444;
}


-- 
Igor Sysoev


From igor at sysoev.ru  Fri May 27 12:00:55 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 12:00:55 +0400
Subject: How to add cookie when proxy_pass via programming?
In-Reply-To: <3dc5bd76bbf522bacd924a5f91ecc262.NginxMailingListEnglish@forum.nginx.org>
References: <3dc5bd76bbf522bacd924a5f91ecc262.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110527080055.GI79260@sysoev.ru>

On Fri, May 27, 2011 at 12:49:30AM -0400, speedfirst wrote:
> Hey, I want to add custom cookie when proxy pass to the upstream in my
> custom module. I want to implement like what this config does:
> 
> proxy_pass http://servers;
> proxy_add_header Cookie "$http_cookie;KEY=VALUE";
> 
> But sinc the VALUE is dynamically generated so I need to add this cookie
> via programming.
> 
> How to do that?

Create a module that defines a veriable and use it in
proxy_set_header  Cookie  "$http_cookie; key=$var";


-- 
Igor Sysoev


From igor at sysoev.ru  Fri May 27 12:10:50 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 12:10:50 +0400
Subject: SMTP or POP3 Authentication
In-Reply-To: <6C6276B0EF904519A3133AB1D85ED644@joshparker.us>
References: <6C6276B0EF904519A3133AB1D85ED644@joshparker.us>
Message-ID: <20110527081050.GA95514@sysoev.ru>

On Fri, May 27, 2011 at 12:17:10AM -0400, Parker, Joshua wrote:
> I am trying to use smtp connection for an email marketing software and would like to use smtp or pop3, but I am not sure where to start. I have mailauth.pm in /root/auth/ but I am not sure if that is write. If the previous is correct, is this also correct in nginx.conf:
> 
> mail { auth_http 127.0.0.1:80/auth; pop3_capabilities "TOP" "USER";
>  smtp_auth plain login; server { listen 110; protocol pop3; proxy on; } }
> 
> If none of the above is correct, can you point me in the write direction or documentation on how to accomplish this? Thanks.

nginx can not work as SMTP or POP3 server, it can only proxy connections
to real mail servers.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May 27 12:14:49 2011
From: nginx-forum at nginx.us (pk899)
Date: Fri, 27 May 2011 04:14:49 -0400
Subject: Newbie questions about nginx (moving from apache)
Message-ID: <6f56cffc73aa96c01438b44bd1386150.NginxMailingListEnglish@forum.nginx.org>

Hi 

sorry for english. 

i am on centos 64 bit server with cpanel and whm. 

i have apache 2.2.18 and there is heavy duty use of the following: 

1. mod rewrite 
2. mod geoip
3. mod security (with rules available in 2.6.0 for google safe browsing
checks etc)  
4. mod evasive 
5. php5 functionality with memcached + eaccelerator 

for #1, i see from threads on this phorum that some rewriting
functionality is possible. i also see that in 1.0.3 the #2 of geoip is
possible. 

but i am not sure about 3, 4 and 5. nginx does not support DDOS
protection right? also, will all the php5 modules work with pecl such as
memcached etc? 

in trying to set up nginx, i have problem getting php to work. mod fcgi
-- is there simple instruction about how to do, and how is the
performance when compared to apache+eaccelerator which opcode caches? 

because of this functionality, i don't want to move entire to nginx. 

so i am trying this: nginx run on localhost on port 81. apache is main
front server on port 80. all static content from images folder need to
pass to nginx. i am trying this code in apache: 


ProxyRequests       Off
ProxyPreserveHost   On
ProxyPass           /mysite/images    http://0.0.0.0:81/
ProxyPassReverse    /           http://0.0.0.0:81/


But this is not working. 

Nginx root folder is "/mysite/images". And when I type in browser
http://MYDOMAIN.COM:81 then I can see the nginx working fine. the images
are there. 

any idea about how i can make this setup work? 

or ideally, over time i want to move entirely to nginx. so would
appreciate some experienced thought or guidance. 

thanks

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201449,201449#msg-201449



From igor at sysoev.ru  Fri May 27 12:24:14 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 12:24:14 +0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <6f56cffc73aa96c01438b44bd1386150.NginxMailingListEnglish@forum.nginx.org>
References: <6f56cffc73aa96c01438b44bd1386150.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110527082414.GB95514@sysoev.ru>

On Fri, May 27, 2011 at 04:14:49AM -0400, pk899 wrote:
> Hi 
> 
> sorry for english. 
> 
> i am on centos 64 bit server with cpanel and whm. 
> 
> i have apache 2.2.18 and there is heavy duty use of the following: 
> 
> 1. mod rewrite 
> 2. mod geoip
> 3. mod security (with rules available in 2.6.0 for google safe browsing
> checks etc)  
> 4. mod evasive 
> 5. php5 functionality with memcached + eaccelerator 
> 
> for #1, i see from threads on this phorum that some rewriting
> functionality is possible. i also see that in 1.0.3 the #2 of geoip is
> possible. 
> 
> but i am not sure about 3, 4 and 5. nginx does not support DDOS
> protection right? also, will all the php5 modules work with pecl such as
> memcached etc? 
> 
> in trying to set up nginx, i have problem getting php to work. mod fcgi
> -- is there simple instruction about how to do, and how is the
> performance when compared to apache+eaccelerator which opcode caches? 
> 
> because of this functionality, i don't want to move entire to nginx. 
> 
> so i am trying this: nginx run on localhost on port 81. apache is main
> front server on port 80. all static content from images folder need to
> pass to nginx. i am trying this code in apache: 
> 
> 
> ProxyRequests       Off
> ProxyPreserveHost   On
> ProxyPass           /mysite/images    http://0.0.0.0:81/
> ProxyPassReverse    /           http://0.0.0.0:81/
> 
> 
> But this is not working. 
> 
> Nginx root folder is "/mysite/images". And when I type in browser
> http://MYDOMAIN.COM:81 then I can see the nginx working fine. the images
> are there. 
> 
> any idea about how i can make this setup work? 
> 
> or ideally, over time i want to move entirely to nginx. so would
> appreciate some experienced thought or guidance. 

You should use 127.0.0.1:81 instead of 0.0.0.0:81.
However, it's better to set nginx before Apache and all except
"/mysite/images" to the Apache. 
nginx has not mod_security functionality.
There is some functionality similar to mod_evasive in limit_req.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May 27 12:28:10 2011
From: nginx-forum at nginx.us (pk899)
Date: Fri, 27 May 2011 04:28:10 -0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <20110527082414.GB95514@sysoev.ru>
References: <20110527082414.GB95514@sysoev.ru>
Message-ID: <7222f6608bdfa123403f17dbd24ae2ca.NginxMailingListEnglish@forum.nginx.org>

thanks igor. wow, so fast reply. 

how can i setup nginx "before apache"? will i have to change apache
port? 

and what will the syntax look like. is there any simple guide? i have a
high traffic site so dont want to break the functionality. 

thanks!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201449,201454#msg-201454



From igor at sysoev.ru  Fri May 27 12:34:14 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 12:34:14 +0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <7222f6608bdfa123403f17dbd24ae2ca.NginxMailingListEnglish@forum.nginx.org>
References: <20110527082414.GB95514@sysoev.ru>
	<7222f6608bdfa123403f17dbd24ae2ca.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110527083414.GC95514@sysoev.ru>

On Fri, May 27, 2011 at 04:28:10AM -0400, pk899 wrote:
> thanks igor. wow, so fast reply. 
> 
> how can i setup nginx "before apache"? will i have to change apache
> port? 
> 
> and what will the syntax look like. is there any simple guide? i have a
> high traffic site so dont want to break the functionality. 

You should test nginx setup on 81 port passing request to Apache on 80:

server {
    listen  81;
    server_name  your.domain.com;

    root  /path/to/static;

    location / {
        proxy_pass        http://127.0.0.1;
        proxy_set_header  Host  $host;
        proxy_set_header  X-Forwarded-For  $proxy_add_x_forwarded_for;
    }

    location /mysite/images/ {
    }

}

You may need to add mod_rpaf to Apache to process client IPs passed
by nginx in X-Forwarded-For.


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May 27 12:37:43 2011
From: nginx-forum at nginx.us (Allison88)
Date: Fri, 27 May 2011 04:37:43 -0400
Subject: Wether Borbon will appear on the court depends on his test on Friday
Message-ID: <23a0ce9c48f19d7db4589c84fdd6706f.NginxMailingListEnglish@forum.nginx.org>

Julio Borbon will test his inflamed hamstring Friday by running the
bases. Borbon has only run at about 60-70 percent this week. If the
Friday test goes well, he could begin a rehab assignment over the
weekend. Manager Ron Washington said Borbon would not return Sunday, the
first day he's eligible. More relate info on
[b][url=http://www.texas-rangers-shop.com]Texas Rangers
jerseys[/url][/b] zone

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201458,201458#msg-201458



From Dalibor.Jotanovic at herold.at  Fri May 27 13:30:17 2011
From: Dalibor.Jotanovic at herold.at (Dalibor Jotanovic)
Date: Fri, 27 May 2011 11:30:17 +0200
Subject: =?UTF-8?Q?NGINX_1=2E0=2E3_compile_error=3A_src/core/ngx=5Fcrypt=2Ec=3A_In_?=
	=?UTF-8?Q?function_=E2=80=98ngx=5Fcrypt=5Fapr1=E2=80=99?=
References: <4DDF83280200005E000347A4@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
Message-ID: <4DDF8B490200005E000347B9@mx.herold.at>

Hi,
 
i'm having problems compiling the latest stable source 1.0.3.
 
How to reproduce:
 
cd <sourcedirectory>
./configure
make
 
make output is attached in log.txt
 
I do not have any problems compiling stable source 1.0.2
 
Is suspect this issue beeing a bug. Can somebody confirm this?
 
Thanks in advance!
 
Cheers,
DJ
 

HEROLD Business Data GmbH
Guntramsdorfer Stra?e 105
A-2340 M?dling

FN 233171z
Landesgericht Wiener Neustadt 
  
Besuchen Sie uns online und mobil www.herold.at!

Weitere Informationen zu unseren Produkten finden Sie unter:
http://mobile.herold.at
http://daten.herold.at

Werden Sie Fan von HEROLD auf Facebook!


Bitte beachten Sie auch unsere Tochter- und Partner-Unternehmen:
www.clearsense.at
www.tupalo.com
www.aboutmedia.at
www.love.at
www.reise-hero.at
www.urlauburlaub.at
 
Diese E-Mail kann vertrauliche und/oder rechtlich gesch?tzte
Informationen enthalten. Wenn Sie nicht der richtige Adressat sind oder
diese E-Mail irrt?mlich erhalten haben, informieren Sie bitte sofort den
Absender und vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie
die unbefugte Weitergabe dieser Mail sind nicht gestattet. 

This e-mail may contain confidential and/or privileged information. If
you are not the intended recipient (or have received this e-mail in
error) please notify the sender immediately and delete this e-mail. Any
unauthorized copying, disclosure or distribution of the material in this
e-mail is strictly forbidden. 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/775c42ff/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1993 bytes
Desc: CompuServe GIF graphic
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/775c42ff/attachment-0002.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1096 bytes
Desc: CompuServe GIF graphic
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/775c42ff/attachment-0003.gif>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: log.txt
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/775c42ff/attachment-0001.txt>

From nginx-forum at nginx.us  Fri May 27 13:32:11 2011
From: nginx-forum at nginx.us (speedfirst)
Date: Fri, 27 May 2011 05:32:11 -0400
Subject: How to add cookie when proxy_pass via programming?
In-Reply-To: <3dc5bd76bbf522bacd924a5f91ecc262.NginxMailingListEnglish@forum.nginx.org>
References: <3dc5bd76bbf522bacd924a5f91ecc262.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <e7a4db4a3bdd894abab8b009f694b779.NginxMailingListEnglish@forum.nginx.org>

Thanks. And I want to know the time when "proxy_set_header" is executed,
because the my cookie value is related to the chosen upstream server. So
will proxy_set_header happen before upstream peer initialization or
after that?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201413,201474#msg-201474



From igor at sysoev.ru  Fri May 27 13:44:16 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 13:44:16 +0400
Subject: How to add cookie when proxy_pass via programming?
In-Reply-To: <e7a4db4a3bdd894abab8b009f694b779.NginxMailingListEnglish@forum.nginx.org>
References: <3dc5bd76bbf522bacd924a5f91ecc262.NginxMailingListEnglish@forum.nginx.org>
	<e7a4db4a3bdd894abab8b009f694b779.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110527094416.GD95514@sysoev.ru>

On Fri, May 27, 2011 at 05:32:11AM -0400, speedfirst wrote:
> Thanks. And I want to know the time when "proxy_set_header" is executed,
> because the my cookie value is related to the chosen upstream server. So
> will proxy_set_header happen before upstream peer initialization or
> after that?

Before, and this value is used for all upstreams.


-- 
Igor Sysoev


From igor at sysoev.ru  Fri May 27 13:57:45 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 13:57:45 +0400
Subject: =?UTF-8?Q?Re=3A_NGINX_1=2E0=2E3_compile_error=3A_src/core/ngx=5Fcrypt=2Ec?=
	=?UTF-8?Q?=3A_In_function_=E2=80=98ngx=5Fcrypt=5Fapr1=E2=80=99?=
In-Reply-To: <4DDF8B490200005E000347B9@mx.herold.at>
References: <4DDF83280200005E000347A4@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
Message-ID: <20110527095745.GE95514@sysoev.ru>

On Fri, May 27, 2011 at 11:30:17AM +0200, Dalibor Jotanovic wrote:
> Hi,
>  
> i'm having problems compiling the latest stable source 1.0.3.
>  
> How to reproduce:
>  
> cd <sourcedirectory>
> ./configure
> make
>  
> make output is attached in log.txt
>  
> I do not have any problems compiling stable source 1.0.2
>  
> Is suspect this issue beeing a bug. Can somebody confirm this?

> gcc -c -pipe  -O -W -Wall -Wpointer-arith -Wno-unused-parameter -Wunused-function -Wunused-variable -Wunused-value -Werror -g  -I src/core -I 
> src/event -I src/event/modules -I src/os/unix -I objs \
> 		-o objs/src/core/ngx_crypt.o \
> 		src/core/ngx_crypt.c
> cc1: warnings being treated as errors
> src/core/ngx_crypt.c: In function ???ngx_crypt_apr1???:
> src/core/ngx_crypt.c:76: warning: pointer targets in passing argument 2 of ???ngx_md5_update??? differ in signedness

Thank you for the report.
The attached patch should fix the issue.


-- 
Igor Sysoev
-------------- next part --------------
Index: src/core/ngx_md5.c
===================================================================
--- src/core/ngx_md5.c	(revision 3927)
+++ src/core/ngx_md5.c	(working copy)
@@ -32,7 +32,7 @@
 
 
 void
-ngx_md5_update(ngx_md5_t *ctx, const u_char *data, size_t size)
+ngx_md5_update(ngx_md5_t *ctx, const void *data, size_t size)
 {
     size_t  used, free;
 
@@ -47,8 +47,7 @@
             return;
         }
 
-        ngx_memcpy(&ctx->buffer[used], data, free);
-        data = (u_char *)data + free;
+        data = ngx_cpymem(&ctx->buffer[used], data, free);
         size -= free;
         (void) ngx_md5_body(ctx, ctx->buffer, 64);
     }
Index: src/core/ngx_md5.h
===================================================================
--- src/core/ngx_md5.h	(revision 3927)
+++ src/core/ngx_md5.h	(working copy)
@@ -50,7 +50,7 @@
 
 
 void ngx_md5_init(ngx_md5_t *ctx);
-void ngx_md5_update(ngx_md5_t *ctx, const u_char *data, size_t size);
+void ngx_md5_update(ngx_md5_t *ctx, const void *data, size_t size);
 void ngx_md5_final(u_char result[16], ngx_md5_t *ctx);
 
 

From Dalibor.Jotanovic at herold.at  Fri May 27 14:38:36 2011
From: Dalibor.Jotanovic at herold.at (Dalibor Jotanovic)
Date: Fri, 27 May 2011 12:38:36 +0200
Subject: =?UTF-8?Q?Antw=3A_Re=3A_NGINX_1=2E0=2E3_compile_error=3A_src/core/ngx=5Fcr?=
	=?UTF-8?Q?ypt=2Ec=3A_In_function_=E2=80=98ngx=5Fcrypt=5Fapr1=E2=80=99?=
In-Reply-To: <20110527095745.GE95514@sysoev.ru>
References: <4DDF83280200005E000347A4@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
	<20110527095745.GE95514@sysoev.ru>
Message-ID: <4DDF9B4C0200005E0003480A@mx.herold.at>

Hi,
 
that patch fixed the issue. Many thanks for the very fast support!
 
Regards,
DJ

>>> Igor Sysoev <igor at sysoev.ru> 5/27/2011 11:57 >>>
On Fri, May 27, 2011 at 11:30:17AM +0200, Dalibor Jotanovic wrote:
> Hi,
>  
> i'm having problems compiling the latest stable source 1.0.3.
>  
> How to reproduce:
>  
> cd <sourcedirectory>
> ./configure
> make
>  
> make output is attached in log.txt
>  
> I do not have any problems compiling stable source 1.0.2
>  
> Is suspect this issue beeing a bug. Can somebody confirm this?

> gcc -c -pipe  -O -W -Wall -Wpointer-arith -Wno-unused-parameter
-Wunused-function -Wunused-variable -Wunused-value -Werror -g  -I
src/core -I 
> src/event -I src/event/modules -I src/os/unix -I objs \
> -o objs/src/core/ngx_crypt.o \
> src/core/ngx_crypt.c
> cc1: warnings being treated as errors
> src/core/ngx_crypt.c: In function ???ngx_crypt_apr1???:
> src/core/ngx_crypt.c:76: warning: pointer targets in passing argument
2 of ???ngx_md5_update??? differ in signedness

Thank you for the report.
The attached patch should fix the issue.


-- 
Igor Sysoev

HEROLD Business Data GmbH
Guntramsdorfer Stra?e 105
A-2340 M?dling

FN 233171z
Landesgericht Wiener Neustadt 
  
Besuchen Sie uns online und mobil www.herold.at!

Weitere Informationen zu unseren Produkten finden Sie unter:
http://mobile.herold.at
http://daten.herold.at

Werden Sie Fan von HEROLD auf Facebook!


Bitte beachten Sie auch unsere Tochter- und Partner-Unternehmen:
www.clearsense.at
www.tupalo.com
www.aboutmedia.at
www.love.at
www.reise-hero.at
www.urlauburlaub.at
 
Diese E-Mail kann vertrauliche und/oder rechtlich gesch?tzte
Informationen enthalten. Wenn Sie nicht der richtige Adressat sind oder
diese E-Mail irrt?mlich erhalten haben, informieren Sie bitte sofort den
Absender und vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie
die unbefugte Weitergabe dieser Mail sind nicht gestattet. 

This e-mail may contain confidential and/or privileged information. If
you are not the intended recipient (or have received this e-mail in
error) please notify the sender immediately and delete this e-mail. Any
unauthorized copying, disclosure or distribution of the material in this
e-mail is strictly forbidden. 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/86504ad4/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1993 bytes
Desc: CompuServe GIF graphic
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/86504ad4/attachment.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1096 bytes
Desc: CompuServe GIF graphic
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/86504ad4/attachment-0001.gif>

From Dalibor.Jotanovic at herold.at  Fri May 27 14:43:16 2011
From: Dalibor.Jotanovic at herold.at (Dalibor Jotanovic)
Date: Fri, 27 May 2011 12:43:16 +0200
Subject: NGINX loadbalancing ORACLE forms backends?
Message-ID: <4DDF9C640200005E0003480F@mx.herold.at>

Hi,
 
does anybody have a working setup with NGINX loadbalancing ORACLE FORMS
backends?
I'm interested in setting up an environment with JSESSION cookie based
sticky sessions.
 
Regards,
DJ
 

HEROLD Business Data GmbH
Guntramsdorfer Stra?e 105
A-2340 M?dling

FN 233171z
Landesgericht Wiener Neustadt 
  
Besuchen Sie uns online und mobil www.herold.at!

Weitere Informationen zu unseren Produkten finden Sie unter:
http://mobile.herold.at
http://daten.herold.at

Werden Sie Fan von HEROLD auf Facebook!


Bitte beachten Sie auch unsere Tochter- und Partner-Unternehmen:
www.clearsense.at
www.tupalo.com
www.aboutmedia.at
www.love.at
www.reise-hero.at
www.urlauburlaub.at
 
Diese E-Mail kann vertrauliche und/oder rechtlich gesch?tzte
Informationen enthalten. Wenn Sie nicht der richtige Adressat sind oder
diese E-Mail irrt?mlich erhalten haben, informieren Sie bitte sofort den
Absender und vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie
die unbefugte Weitergabe dieser Mail sind nicht gestattet. 

This e-mail may contain confidential and/or privileged information. If
you are not the intended recipient (or have received this e-mail in
error) please notify the sender immediately and delete this e-mail. Any
unauthorized copying, disclosure or distribution of the material in this
e-mail is strictly forbidden. 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/b3c046dd/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1993 bytes
Desc: CompuServe GIF graphic
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/b3c046dd/attachment-0002.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1096 bytes
Desc: CompuServe GIF graphic
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/b3c046dd/attachment-0003.gif>

From jerome at loyet.net  Fri May 27 14:52:09 2011
From: jerome at loyet.net (=?ISO-8859-1?B?Suly9G1lIExveWV0?=)
Date: Fri, 27 May 2011 12:52:09 +0200
Subject: NGINX loadbalancing ORACLE forms backends?
In-Reply-To: <4DDF9C640200005E0003480F@mx.herold.at>
References: <4DDF9C640200005E0003480F@mx.herold.at>
Message-ID: <BANLkTi=us=EwWDtsKehK-LVQH7LkXkWH5w@mail.gmail.com>

Hi

have a look on this module I've made:
http://code.google.com/p/nginx-sticky-module/

I'm using it in different environments (PHP, tomcat, jboss, ...). It can
feet your need. Moreover it's independant from the type of backend you're
using.

hope this help

++ Jerome

2011/5/27 Dalibor Jotanovic <Dalibor.Jotanovic at herold.at>

>  Hi,
>
> does anybody have a working setup with NGINX loadbalancing ORACLE FORMS
> backends?
> I'm interested in setting up an environment with JSESSION cookie based
> sticky sessions.
>
> Regards,
> DJ
>
>
>   <http://www.herold.at>
> <http://www.herold.at>
>
> HEROLD Business Data GmbH
> Guntramsdorfer Stra?e 105
> A-2340 M?dling
>
> FN 233171z
> Landesgericht Wiener Neustadt
>
> Besuchen Sie uns online und mobil auf *www.herold.at*<http://www.herold.at>
> !
>
> Weitere Informationen zu unseren Produkten finden Sie unter:
> *http://mobile.herold.at* <http://mobile.herold.at>
> *http://daten.herold.at * <http://daten.herold.at>
>
> <http://www.facebook.com/derherold>Werden Sie Fan von HEROLD auf Facebook!
>
>
> Bitte beachten Sie auch unsere Tochter- und Partner-Unternehmen:
> *www.clearsense.at * <http://www.clearsense.at>
> *www.tupalo.com * <http://www.tupalo.com>
> *www.aboutmedia.at* <http://www.aboutmedia.at>* <http://www.tupalo.com>*
> *www.love.at* <http://www.love.at>* <http://www.tupalo.com>*
> *www.reise-hero.at* <http://www.reise-hero.at>* <http://www.tupalo.com>*
> *www.urlauburlaub.at* <http://www.urlauburlaub.at>*
> <http://www.tupalo.com>*
>  <http://www.tupalo.com>
> Diese E-Mail kann vertrauliche und/oder rechtlich gesch?tzte Informationen
> enthalten. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail
> irrt?mlich erhalten haben, informieren Sie bitte sofort den Absender und
> vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie die unbefugte
> Weitergabe dieser Mail sind nicht gestattet.
>
> This e-mail may contain confidential and/or privileged information. If you
> are not the intended recipient (or have received this e-mail in error)
> please notify the sender immediately and delete this e-mail. Any
> unauthorized copying, disclosure or distribution of the material in this
> e-mail is strictly forbidden.
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/11433ab5/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1993 bytes
Desc: not available
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/11433ab5/attachment.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1096 bytes
Desc: not available
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/11433ab5/attachment-0001.gif>

From nbubingo at gmail.com  Fri May 27 15:20:31 2011
From: nbubingo at gmail.com (=?GB2312?B?0qbOsLHz?=)
Date: Fri, 27 May 2011 19:20:31 +0800
Subject: NGINX loadbalancing ORACLE forms backends?
In-Reply-To: <4DDF9C640200005E0003480F@mx.herold.at>
References: <4DDF9C640200005E0003480F@mx.herold.at>
Message-ID: <BANLkTi=m1zbSLoP=3S-Q0gPxUbBJ39_P+w@mail.gmail.com>

See also this module: http://code.google.com/p/nginx-upstream-jvm-route/

<http://code.google.com/p/nginx-upstream-jvm-route/>This module achieves
session stickiness with the session cookie. If the session is not in the
cookie or URL, the module will be a normal Round-Robin upstream module.

2011/5/27 Dalibor Jotanovic <Dalibor.Jotanovic at herold.at>

>  Hi,
>
> does anybody have a working setup with NGINX loadbalancing ORACLE FORMS
> backends?
> I'm interested in setting up an environment with JSESSION cookie based
> sticky sessions.
>
> Regards,
> DJ
>
>
>   <http://www.herold.at>
> <http://www.herold.at>
>
> HEROLD Business Data GmbH
> Guntramsdorfer Stra?e 105
> A-2340 M?dling
>
> FN 233171z
> Landesgericht Wiener Neustadt
>
> Besuchen Sie uns online und mobil auf *www.herold.at*<http://www.herold.at>
> !
>
> Weitere Informationen zu unseren Produkten finden Sie unter:
> *http://mobile.herold.at* <http://mobile.herold.at>
> *http://daten.herold.at * <http://daten.herold.at>
>
> <http://www.facebook.com/derherold>Werden Sie Fan von HEROLD auf Facebook!
>
>
> Bitte beachten Sie auch unsere Tochter- und Partner-Unternehmen:
> *www.clearsense.at * <http://www.clearsense.at>
> *www.tupalo.com * <http://www.tupalo.com>
> *www.aboutmedia.at* <http://www.aboutmedia.at>* <http://www.tupalo.com>*
> *www.love.at* <http://www.love.at>* <http://www.tupalo.com>*
> *www.reise-hero.at* <http://www.reise-hero.at>* <http://www.tupalo.com>*
> *www.urlauburlaub.at* <http://www.urlauburlaub.at>*
> <http://www.tupalo.com>*
>  <http://www.tupalo.com>
> Diese E-Mail kann vertrauliche und/oder rechtlich gesch?tzte Informationen
> enthalten. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail
> irrt?mlich erhalten haben, informieren Sie bitte sofort den Absender und
> vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie die unbefugte
> Weitergabe dieser Mail sind nicht gestattet.
>
> This e-mail may contain confidential and/or privileged information. If you
> are not the intended recipient (or have received this e-mail in error)
> please notify the sender immediately and delete this e-mail. Any
> unauthorized copying, disclosure or distribution of the material in this
> e-mail is strictly forbidden.
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/35df8974/attachment.html>

From Pekka.Panula at sofor.fi  Fri May 27 16:02:17 2011
From: Pekka.Panula at sofor.fi (Pekka.Panula at sofor.fi)
Date: Fri, 27 May 2011 15:02:17 +0300
Subject: TLS 1.1 and TLS 1.2 possible?
Message-ID: <OF674B0C14.77CF01CA-ONC225789D.004200E2-C225789D.0042213E@sofor.fi>

I am looking for info is it possible to compile nginx SSL routines with 
TLS 1.1 and 1.2 support? Is openssl only possible backend for nginx?

Terveisin/Regards,
   Pekka Panula, Sofor Oy - Jatkuvat palvelut

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110527/9cbae191/attachment.html>

From igor at sysoev.ru  Fri May 27 16:07:14 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 16:07:14 +0400
Subject: TLS 1.1 and TLS 1.2 possible?
In-Reply-To: <OF674B0C14.77CF01CA-ONC225789D.004200E2-C225789D.0042213E@sofor.fi>
References: <OF674B0C14.77CF01CA-ONC225789D.004200E2-C225789D.0042213E@sofor.fi>
Message-ID: <20110527120714.GJ95514@sysoev.ru>

On Fri, May 27, 2011 at 03:02:17PM +0300, Pekka.Panula at sofor.fi wrote:
> I am looking for info is it possible to compile nginx SSL routines with 
> TLS 1.1 and 1.2 support? Is openssl only possible backend for nginx?

Currently nginx supports only OpenSSL and relies OpenSSL TLS suuport.


-- 
Igor Sysoev


From mdounin at mdounin.ru  Fri May 27 16:07:47 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 27 May 2011 16:07:47 +0400
Subject: =?UTF-8?Q?Re=3A_NGINX_1=2E0=2E3_compile_error=3A_src/core/ngx=5Fcrypt=2Ec?=
	=?UTF-8?Q?=3A_In_function_=E2=80=98ngx=5Fcrypt=5Fapr1=E2=80=99?=
In-Reply-To: <20110527095745.GE95514@sysoev.ru>
References: <4DDF83280200005E000347A4@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
	<20110527095745.GE95514@sysoev.ru>
Message-ID: <20110527120747.GH42265@mdounin.ru>

Hello!

On Fri, May 27, 2011 at 01:57:45PM +0400, Igor Sysoev wrote:

> On Fri, May 27, 2011 at 11:30:17AM +0200, Dalibor Jotanovic wrote:
> > Hi,
> >  
> > i'm having problems compiling the latest stable source 1.0.3.
> >  
> > How to reproduce:
> >  
> > cd <sourcedirectory>
> > ./configure
> > make
> >  
> > make output is attached in log.txt
> >  
> > I do not have any problems compiling stable source 1.0.2
> >  
> > Is suspect this issue beeing a bug. Can somebody confirm this?
> 
> > gcc -c -pipe  -O -W -Wall -Wpointer-arith -Wno-unused-parameter -Wunused-function -Wunused-variable -Wunused-value -Werror -g  -I src/core -I 
> > src/event -I src/event/modules -I src/os/unix -I objs \
> > 		-o objs/src/core/ngx_crypt.o \
> > 		src/core/ngx_crypt.c
> > cc1: warnings being treated as errors
> > src/core/ngx_crypt.c: In function ???ngx_crypt_apr1???:
> > src/core/ngx_crypt.c:76: warning: pointer targets in passing argument 2 of ???ngx_md5_update??? differ in signedness
> 
> Thank you for the report.
> The attached patch should fix the issue.
> 
> 
> -- 
> Igor Sysoev

> Index: src/core/ngx_md5.c
> ===================================================================
> --- src/core/ngx_md5.c	(revision 3927)
> +++ src/core/ngx_md5.c	(working copy)
> @@ -32,7 +32,7 @@
>  
>  
>  void
> -ngx_md5_update(ngx_md5_t *ctx, const u_char *data, size_t size)
> +ngx_md5_update(ngx_md5_t *ctx, const void *data, size_t size)

$ grep MD5Update /usr/include/sys/md5.h
void   MD5Update (MD5_CTX *, const unsigned char *, unsigned int);

So this is required anyway:

diff --git a/src/core/ngx_crypt.c b/src/core/ngx_crypt.c
--- a/src/core/ngx_crypt.c
+++ b/src/core/ngx_crypt.c
@@ -74,7 +74,7 @@ ngx_crypt_apr1(ngx_pool_t *pool, u_char

     ngx_md5_init(&md5);
     ngx_md5_update(&md5, key, keylen);
-    ngx_md5_update(&md5, "$apr1$", sizeof("$apr1$") - 1);
+    ngx_md5_update(&md5, (u_char *) "$apr1$", sizeof("$apr1$") - 1);
     ngx_md5_update(&md5, salt, saltlen);

     ngx_md5_init(&ctx1);

Sorry for the breakage.  BTW, it's probably good idea to finally 
branch stable-1.0 and start releasing 1.1.*.

Maxim Dounin


From igor at sysoev.ru  Fri May 27 16:18:30 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 16:18:30 +0400
Subject: =?UTF-8?Q?Re=3A_NGINX_1=2E0=2E3_compile_error=3A_src/core/ngx=5Fcrypt=2Ec?=
	=?UTF-8?Q?=3A_In_function_=E2=80=98ngx=5Fcrypt=5Fapr1=E2=80=99?=
In-Reply-To: <20110527120747.GH42265@mdounin.ru>
References: <4DDF83280200005E000347A4@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
	<20110527095745.GE95514@sysoev.ru>
	<20110527120747.GH42265@mdounin.ru>
Message-ID: <20110527121830.GK95514@sysoev.ru>

On Fri, May 27, 2011 at 04:07:47PM +0400, Maxim Dounin wrote:

> > Index: src/core/ngx_md5.c
> > ===================================================================
> > --- src/core/ngx_md5.c	(revision 3927)
> > +++ src/core/ngx_md5.c	(working copy)
> > @@ -32,7 +32,7 @@
> >  
> >  
> >  void
> > -ngx_md5_update(ngx_md5_t *ctx, const u_char *data, size_t size)
> > +ngx_md5_update(ngx_md5_t *ctx, const void *data, size_t size)
> 
> $ grep MD5Update /usr/include/sys/md5.h
> void   MD5Update (MD5_CTX *, const unsigned char *, unsigned int);

What OS ?

$ man 3 md5 (MacOSX 10.6)
        [ ... ]
        #include <openssl/md5.h>
        [ ... ]
        int MD5_Update(MD5_CTX *c, const void *data,
                         unsigned long len);

$ man 3 md5 (FreeBSD 7-8)
     [ ... ]
     #include <md5.h>
     [ ... ]
     MD5Update(MD5_CTX *context, const void *data, unsigned int len);

$ grep MD5Update /usr/include/sys/md5.h 
void   MD5Update (MD5_CTX *, const void *, unsigned int);


-- 
Igor Sysoev


From igor at sysoev.ru  Fri May 27 16:34:07 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 16:34:07 +0400
Subject: =?UTF-8?Q?Re=3A_NGINX_1=2E0=2E3_compile_error=3A_src/core/ngx=5Fcrypt=2Ec?=
	=?UTF-8?Q?=3A_In_function_=E2=80=98ngx=5Fcrypt=5Fapr1=E2=80=99?=
In-Reply-To: <20110527121830.GK95514@sysoev.ru>
References: <4DDF83280200005E000347A4@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
	<20110527095745.GE95514@sysoev.ru>
	<20110527120747.GH42265@mdounin.ru>
	<20110527121830.GK95514@sysoev.ru>
Message-ID: <20110527123407.GL95514@sysoev.ru>

On Fri, May 27, 2011 at 04:18:30PM +0400, Igor Sysoev wrote:
> On Fri, May 27, 2011 at 04:07:47PM +0400, Maxim Dounin wrote:
> 
> > > Index: src/core/ngx_md5.c
> > > ===================================================================
> > > --- src/core/ngx_md5.c	(revision 3927)
> > > +++ src/core/ngx_md5.c	(working copy)
> > > @@ -32,7 +32,7 @@
> > >  
> > >  
> > >  void
> > > -ngx_md5_update(ngx_md5_t *ctx, const u_char *data, size_t size)
> > > +ngx_md5_update(ngx_md5_t *ctx, const void *data, size_t size)
> > 
> > $ grep MD5Update /usr/include/sys/md5.h
> > void   MD5Update (MD5_CTX *, const unsigned char *, unsigned int);
> 
> What OS ?
> 
> $ man 3 md5 (MacOSX 10.6)
>         [ ... ]
>         #include <openssl/md5.h>
>         [ ... ]
>         int MD5_Update(MD5_CTX *c, const void *data,
>                          unsigned long len);
> 
> $ man 3 md5 (FreeBSD 7-8)
>      [ ... ]
>      #include <md5.h>
>      [ ... ]
>      MD5Update(MD5_CTX *context, const void *data, unsigned int len);
> 
> $ grep MD5Update /usr/include/sys/md5.h 
> void   MD5Update (MD5_CTX *, const void *, unsigned int);

It seems you looked at FreeBSD prior to 7.0-CURRENT,
this has been changed on 17.01.2006.


-- 
Igor Sysoev


From mdounin at mdounin.ru  Fri May 27 16:36:22 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 27 May 2011 16:36:22 +0400
Subject: =?UTF-8?Q?Re=3A_NGINX_1=2E0=2E3_compile_error=3A_src/core/ngx=5Fcrypt=2Ec?=
	=?UTF-8?Q?=3A_In_function_=E2=80=98ngx=5Fcrypt=5Fapr1=E2=80=99?=
In-Reply-To: <20110527121830.GK95514@sysoev.ru>
References: <4DDF83280200005E000347A4@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
	<4DDF8B490200005E000347B9@mx.herold.at>
	<20110527095745.GE95514@sysoev.ru>
	<20110527120747.GH42265@mdounin.ru>
	<20110527121830.GK95514@sysoev.ru>
Message-ID: <20110527123621.GI42265@mdounin.ru>

Hello!

On Fri, May 27, 2011 at 04:18:30PM +0400, Igor Sysoev wrote:

> On Fri, May 27, 2011 at 04:07:47PM +0400, Maxim Dounin wrote:
> 
> > > Index: src/core/ngx_md5.c
> > > ===================================================================
> > > --- src/core/ngx_md5.c	(revision 3927)
> > > +++ src/core/ngx_md5.c	(working copy)
> > > @@ -32,7 +32,7 @@
> > >  
> > >  
> > >  void
> > > -ngx_md5_update(ngx_md5_t *ctx, const u_char *data, size_t size)
> > > +ngx_md5_update(ngx_md5_t *ctx, const void *data, size_t size)
> > 
> > $ grep MD5Update /usr/include/sys/md5.h
> > void   MD5Update (MD5_CTX *, const unsigned char *, unsigned int);
> 
> What OS ?

FreeBSD 6.

This was changed in 7+, see
http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/sys/md5.h#rev1.19

Maxim Dounin


From nginx-forum at nginx.us  Fri May 27 18:59:41 2011
From: nginx-forum at nginx.us (kheraud)
Date: Fri, 27 May 2011 10:59:41 -0400
Subject: Reverse proxy to a tomcat service (folder name)
Message-ID: <66af54152e33276572536226ae532b7b.NginxMailingListEnglish@forum.nginx.org>

Hello,

I am breaking my head against the wall since his morning on this issue
maybe you can help me.

I have a single physical server hosting tomcat WARs and also other
standalone services. I use subdomains and Nginx to route to the
requested service.
Tomcat append to URIs the name of the deployed WAR. For example, if I
deploy myApp.war on my tomcat server, I can access it through
http://mydomain.com:8080/myApp/.
What I want is to acces my service through http://myApp.domain.com and
navigate my service AND whitout having http://myApp.domain.com/myApp/ !

I wrote this conf :

server {
        server_name myapp.domain.com;
        root /var/lib/tomcat7/webapps/myapp;
        access_log /var/log/nginx/myapp.domain.com_access.log;
        error_log /var/log/nginx/myapp.domain.com_error.log;

        location / {
                proxy_pass              http://localhost:8080/myapp;
                proxy_redirect          off;
                proxy_set_header        Host            $host;
                proxy_set_header        X-Real-IP       $remote_addr;
                proxy_set_header        X-Forwarded-For
$proxy_add_x_forwarded_for;
        }

        location /myapp {
                proxy_pass              http://localhost:8080/myapp;
                proxy_redirect          off;
                proxy_set_header        Host            $host;
                proxy_set_header        X-Real-IP       $remote_addr;
                proxy_set_header        X-Forwarded-For
$proxy_add_x_forwarded_for;
        }
}

If I access http://myapp.domain.com/, my browser rewrites this url to
http://myapp.domain.com/myapp/. I would like to avoid the  /myapp/
repeated.

Do you have hints or comments to help me ?

Thanks a lot

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201561,201561#msg-201561



From igor at sysoev.ru  Fri May 27 19:11:08 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 19:11:08 +0400
Subject: Reverse proxy to a tomcat service (folder name)
In-Reply-To: <66af54152e33276572536226ae532b7b.NginxMailingListEnglish@forum.nginx.org>
References: <66af54152e33276572536226ae532b7b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110527151108.GA46101@sysoev.ru>

On Fri, May 27, 2011 at 10:59:41AM -0400, kheraud wrote:
> Hello,
> 
> I am breaking my head against the wall since his morning on this issue
> maybe you can help me.
> 
> I have a single physical server hosting tomcat WARs and also other
> standalone services. I use subdomains and Nginx to route to the
> requested service.
> Tomcat append to URIs the name of the deployed WAR. For example, if I
> deploy myApp.war on my tomcat server, I can access it through
> http://mydomain.com:8080/myApp/.
> What I want is to acces my service through http://myApp.domain.com and
> navigate my service AND whitout having http://myApp.domain.com/myApp/ !
> 
> I wrote this conf :
> 
> server {
>         server_name myapp.domain.com;
>         root /var/lib/tomcat7/webapps/myapp;
>         access_log /var/log/nginx/myapp.domain.com_access.log;
>         error_log /var/log/nginx/myapp.domain.com_error.log;
> 
>         location / {
>                 proxy_pass              http://localhost:8080/myapp;
>                 proxy_redirect          off;
>                 proxy_set_header        Host            $host;
>                 proxy_set_header        X-Real-IP       $remote_addr;
>                 proxy_set_header        X-Forwarded-For
> $proxy_add_x_forwarded_for;
>         }
> 
>         location /myapp {
>                 proxy_pass              http://localhost:8080/myapp;
>                 proxy_redirect          off;
>                 proxy_set_header        Host            $host;
>                 proxy_set_header        X-Real-IP       $remote_addr;
>                 proxy_set_header        X-Forwarded-For
> $proxy_add_x_forwarded_for;
>         }
> }
> 
> If I access http://myapp.domain.com/, my browser rewrites this url to
> http://myapp.domain.com/myapp/. I would like to avoid the  /myapp/
> repeated.
> 
> Do you have hints or comments to help me ?

Probably this:

   location / {
       proxy_pass        http://myapp.domain.com:8080/myapp/;
       proxy_set_header  Host             $host;
       proxy_set_header  X-Real-IP        $remote_addr;
       proxy_set_header  X-Forwarded-For  $proxy_add_x_forwarded_for;
       sub_filter        /myapp/          /;
   }


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May 27 20:19:47 2011
From: nginx-forum at nginx.us (pk899)
Date: Fri, 27 May 2011 12:19:47 -0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <7222f6608bdfa123403f17dbd24ae2ca.NginxMailingListEnglish@forum.nginx.org>
References: <20110527082414.GB95514@sysoev.ru>
	<7222f6608bdfa123403f17dbd24ae2ca.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <ca9cb62a711505d5b16ac082b226c4db.NginxMailingListEnglish@forum.nginx.org>

thanks for this. but before I try all this on a production server,
couple questions: 

1. if this works with NGINX on port 81, can this work in the future on
port 80? So will nginx and apache both be on port 80? I doubt it but
want to confirm. 

2. secondly, do you have a production-ready sample of how to set up for
static serving? in apache for example, we set max-age, expires, gzip,
cache-control etc for static files and this is very handy. i read the
wiki manual for nginx and there's a lot of options in nginx too (quite
impressive actually) but i would love to see a production recommended
sample of commands. 

thanks!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201449,201595#msg-201595



From treomon at gmail.com  Fri May 27 20:24:09 2011
From: treomon at gmail.com (Mon Yang)
Date: Fri, 27 May 2011 09:24:09 -0700
Subject: Web Server Internal IP Address/Internal Network Name Disclosure
	Vulnerability
Message-ID: <BANLkTin0XJ=6GDoFJeJMQAO8VwJZUMs48A@mail.gmail.com>

Hi,

I did QualysGuard scan and it says "Web Server Internal IP
Address/Internal Network Name Disclosure Vulnerability".  It suggest
for Apache, I need to:


For Apache Web Server:
Modify the Apache configuration file as follows:
- Set "ServerName" to a proper FQDN.
or
- Use module mod_rewrite to modify the 3xx error message returned by the server.

So what I need to do in nginx to fix this issue?  Thank you.


From nginx-forum at nginx.us  Fri May 27 20:26:47 2011
From: nginx-forum at nginx.us (kheraud)
Date: Fri, 27 May 2011 12:26:47 -0400
Subject: Reverse proxy to a tomcat service (folder name)
In-Reply-To: <66af54152e33276572536226ae532b7b.NginxMailingListEnglish@forum.nginx.org>
References: <66af54152e33276572536226ae532b7b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <9fd495e26490b071d3cf413155011172.NginxMailingListEnglish@forum.nginx.org>

Thank you for the hint,

It seems to work (for the first page) but in fact it leads to some
errors... You added "/" to "proxy_pass http://localhost:8080/myapp". But
I ommitted it after having seen that it leads to the following mistakes
on the tomcat side :

The requested resource (/dev01-portal//j_spring_security_check) is not
available.

The access log is :
[27/May/2011:18:23:32 +0200] "POST /dev01-portal/j_spring_security_check
HTTP/1.1" 404 441 "http://dev01-portal.appssalon.com/" "Mozilla/5.0

I don't know why but it seems it requests
http://localhost:8080/dev01-portal//j_spring_security_check.

Do you know where it comes from ?

Thanks again for the help

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201561,201598#msg-201598



From igor at sysoev.ru  Fri May 27 20:33:07 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 20:33:07 +0400
Subject: Reverse proxy to a tomcat service (folder name)
In-Reply-To: <9fd495e26490b071d3cf413155011172.NginxMailingListEnglish@forum.nginx.org>
References: <66af54152e33276572536226ae532b7b.NginxMailingListEnglish@forum.nginx.org>
	<9fd495e26490b071d3cf413155011172.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110527163307.GD46101@sysoev.ru>

On Fri, May 27, 2011 at 12:26:47PM -0400, kheraud wrote:
> Thank you for the hint,
> 
> It seems to work (for the first page) but in fact it leads to some
> errors... You added "/" to "proxy_pass http://localhost:8080/myapp". But
> I ommitted it after having seen that it leads to the following mistakes
> on the tomcat side :
> 
> The requested resource (/dev01-portal//j_spring_security_check) is not
> available.
> 
> The access log is :
> [27/May/2011:18:23:32 +0200] "POST /dev01-portal/j_spring_security_check
> HTTP/1.1" 404 441 "http://dev01-portal.appssalon.com/" "Mozilla/5.0
> 
> I don't know why but it seems it requests
> http://localhost:8080/dev01-portal//j_spring_security_check.
> 
> Do you know where it comes from ?

It seems you have two locations:

location / {
location /dev01-portal {

You need just one:

location / {


-- 
Igor Sysoev


From igor at sysoev.ru  Fri May 27 20:43:25 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 20:43:25 +0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <ca9cb62a711505d5b16ac082b226c4db.NginxMailingListEnglish@forum.nginx.org>
References: <20110527082414.GB95514@sysoev.ru>
	<7222f6608bdfa123403f17dbd24ae2ca.NginxMailingListEnglish@forum.nginx.org>
	<ca9cb62a711505d5b16ac082b226c4db.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110527164325.GE46101@sysoev.ru>

On Fri, May 27, 2011 at 12:19:47PM -0400, pk899 wrote:
> thanks for this. but before I try all this on a production server,
> couple questions: 
> 
> 1. if this works with NGINX on port 81, can this work in the future on
> port 80? So will nginx and apache both be on port 80? I doubt it but
> want to confirm. 

Apache may listen on 127.0.0.1:80, while nginx may listen on *:80.

> 2. secondly, do you have a production-ready sample of how to set up for
> static serving? in apache for example, we set max-age, expires, gzip,
> cache-control etc for static files and this is very handy. i read the
> wiki manual for nginx and there's a lot of options in nginx too (quite
> impressive actually) but i would love to see a production recommended
> sample of commands. 

location /static/ {
    root     ...
    expires      30d;
    gzip         on;
    gzip_static  on;
}


-- 
Igor Sysoev


From igor at sysoev.ru  Fri May 27 20:50:54 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Fri, 27 May 2011 20:50:54 +0400
Subject: Web Server Internal IP Address/Internal Network Name Disclosure
	Vulnerability
In-Reply-To: <BANLkTin0XJ=6GDoFJeJMQAO8VwJZUMs48A@mail.gmail.com>
References: <BANLkTin0XJ=6GDoFJeJMQAO8VwJZUMs48A@mail.gmail.com>
Message-ID: <20110527165054.GF46101@sysoev.ru>

On Fri, May 27, 2011 at 09:24:09AM -0700, Mon Yang wrote:
> Hi,
> 
> I did QualysGuard scan and it says "Web Server Internal IP
> Address/Internal Network Name Disclosure Vulnerability".  It suggest
> for Apache, I need to:
> 
> 
> For Apache Web Server:
> Modify the Apache configuration file as follows:
> - Set "ServerName" to a proper FQDN.
> or
> - Use module mod_rewrite to modify the 3xx error message returned by the server.
> 
> So what I need to do in nginx to fix this issue?  Thank you.

If you proxy server via nginx, then probably something like this:

server {
    server_name  FQDN;

    location / {
        proxy_pass      http://backend;
        proxy_redirect  http://FQDN/  http://backend/;
    }
}


-- 
Igor Sysoev


From nginx-forum at nginx.us  Fri May 27 20:59:46 2011
From: nginx-forum at nginx.us (pk899)
Date: Fri, 27 May 2011 12:59:46 -0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <6f56cffc73aa96c01438b44bd1386150.NginxMailingListEnglish@forum.nginx.org>
References: <6f56cffc73aa96c01438b44bd1386150.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <405146017f0e366230d436dba381c983.NginxMailingListEnglish@forum.nginx.org>

thanks. i have reading about nginx a lot on the internet. and it seems
to me that many features of mod_security are available inside nginx. 

i like this post:
http://eng.eelcowesemann.nl/linux-unix/nginx/nginx-blocking/

my questions. 
1. how should i block remote_addr based on IP in a certain external
file? 
2. to block xss or such type of injection attacks -- any best practice
rules? 
3. can i block based on "request_uri"? if some hosts are in it
(parameter) then block it. 
4. can i block based on RBL checks? from spamhaus etc. 

thanks!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201449,201617#msg-201617



From jackdesert556 at gmail.com  Fri May 27 23:38:18 2011
From: jackdesert556 at gmail.com (Jack Desert)
Date: Fri, 27 May 2011 15:38:18 -0400
Subject: Getting All Documentation Under One Roof
Message-ID: <20110527153818.2d53fda3@dime>

Hi,


Can I get a raise of hands who is interested in a more cohesive partnership between these two sites:
  
  nginx.org
  wiki.nginx.org

and what thoughts you have on the best way to do it? 


I'm new to Nginx and this mailing list, but I want to help out with documentation. I spent most of yesterday editing the english wiki at wiki.nginx.org. My goal was to make it as coherent as possible, and add the necessary "Getting Started" links to the various pages since it took me forever to find what I was looking for originally. I also fixed a couple places where if you hit the wrong link it takes you to a page in Russian. 

The next wave I'd like to pursue is to blend the English wiki and the english nginx.org site into a more cohesive whole. For example, there are only a few official documents in English--and a lot of english info on the wiki. Why not have them both in the same place? Or at least clear links from one to the other. 

My goal, of course, is for all the newbies out there who come along behind me to be able to find exactly they're looking for on the first try. 

Jack Desert

Technical Writer / Programmer


From cliff at develix.com  Sat May 28 01:38:54 2011
From: cliff at develix.com (Cliff Wells)
Date: Fri, 27 May 2011 14:38:54 -0700
Subject: Getting All Documentation Under One Roof
In-Reply-To: <20110527153818.2d53fda3@dime>
References: <20110527153818.2d53fda3@dime>
Message-ID: <1306532334.31643.6.camel@portable-evil>

On Fri, 2011-05-27 at 15:38 -0400, Jack Desert wrote:
> Hi,
> 
> 
> Can I get a raise of hands who is interested in a more cohesive partnership between these two sites:
>   
>   nginx.org
>   wiki.nginx.org
> 
> and what thoughts you have on the best way to do it? 

Igor writes the documentation on nginx.org.  wiki.nginx.org is just
that: a wiki, managed by the community.

IMO, the best thing if you are seeking some cohesion would be to make
sure that docs at nginx.org are properly linked from the wiki and that
they don't conflict (some info on wiki is outdated or incomplete).

> I'm new to Nginx and this mailing list, but I want to help out with
> documentation. I spent most of yesterday editing the english wiki at
> wiki.nginx.org. My goal was to make it as coherent as possible, and
> add the necessary "Getting Started" links to the various pages since
> it took me forever to find what I was looking for originally.

Was this due to not knowing the correct search term or something else?

>  I also fixed a couple places where if you hit the wrong link it takes
> you to a page in Russian. 
> 
> The next wave I'd like to pursue is to blend the English wiki and the
> english nginx.org site into a more cohesive whole. For example, there
> are only a few official documents in English--and a lot of english
> info on the wiki. Why not have them both in the same place? Or at
> least clear links from one to the other. 

Because one is the official docs and the other is a community wiki.  I
don't think Igor wants random people editing his documentation, nor do I
expect him to update the wiki (that's our job, as users).  Links from
the wiki to the official docs would be good.

> My goal, of course, is for all the newbies out there who come along
> behind me to be able to find exactly they're looking for on the first
> try. 


Cliff



From ngx.eugaia at gmail.com  Sat May 28 02:19:34 2011
From: ngx.eugaia at gmail.com (Eugaia)
Date: Sat, 28 May 2011 01:19:34 +0300
Subject: Getting All Documentation Under One Roof
In-Reply-To: <1306532334.31643.6.camel@portable-evil>
References: <20110527153818.2d53fda3@dime>
	<1306532334.31643.6.camel@portable-evil>
Message-ID: <4DE02376.4040503@gmail.com>

Hi,
> On Fri, 2011-05-27 at 15:38 -0400, Jack Desert wrote:
>
> Can I get a raise of hands who is interested in a more cohesive partnership between these two sites:
>
>    nginx.org
>    wiki.nginx.org
>
> and what thoughts you have on the best way to do it?
I think that it would be good to have a bot automatically check pages on 
nginx.org, which would add a notification to the wiki.nginx.org site 
when a new page was added or an existing one was edited.  This would 
probably make it easier for those who wish to keep the wiki up-to-date, 
since they wouldn't need to search for what needed checking.  There 
could also be a section on the notification page for wiki editors to say 
that they have looked over the changes and made edits etc, so less 
duplication of checking is done.

If the change notifications could have diff information, or perhaps 
otherwise identify which configuration directives etc had been added or 
changed, this would make the task even easier.

I agree that linking to the original docs would also be useful.

Cheers,

Marcus.


From nginx-forum at nginx.us  Sat May 28 07:54:52 2011
From: nginx-forum at nginx.us (pk899)
Date: Fri, 27 May 2011 23:54:52 -0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <6f56cffc73aa96c01438b44bd1386150.NginxMailingListEnglish@forum.nginx.org>
References: <6f56cffc73aa96c01438b44bd1386150.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <63174a38ac8500e2ac91fc3b3255a650.NginxMailingListEnglish@forum.nginx.org>

Some problems with your recommendation, Igor. 

1. There is no directive like "gzip_static". 

2. The proxy_pass that you recommend above -- this does a full 301
redirect to Apache! 

This is really not the solution I am after. 

So, back to my first question, how do I pass the baton from Apache
(which really needs to be my first front server because of all the
features) only for the static files to nginx? 

Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201449,201728#msg-201728



From treomon at gmail.com  Sat May 28 08:02:08 2011
From: treomon at gmail.com (Mon Yang)
Date: Fri, 27 May 2011 21:02:08 -0700
Subject: Web Server Internal IP Address/Internal Network Name Disclosure
	Vulnerability
In-Reply-To: <20110527165054.GF46101@sysoev.ru>
References: <BANLkTin0XJ=6GDoFJeJMQAO8VwJZUMs48A@mail.gmail.com>
	<20110527165054.GF46101@sysoev.ru>
Message-ID: <BANLkTinXKgKd9f3v7Sze=46y=QnmagxuWg@mail.gmail.com>

Sorry...  FQDN = nginx server itself or the backend mail server?

Or, can you tell me how to change 3xx error message, since it might be
safer so it won't break things?

Thank you.

On Fri, May 27, 2011 at 9:50 AM, Igor Sysoev <igor at sysoev.ru> wrote:
> On Fri, May 27, 2011 at 09:24:09AM -0700, Mon Yang wrote:
>> Hi,
>>
>> I did QualysGuard scan and it says "Web Server Internal IP
>> Address/Internal Network Name Disclosure Vulnerability". ?It suggest
>> for Apache, I need to:
>>
>>
>> For Apache Web Server:
>> Modify the Apache configuration file as follows:
>> - Set "ServerName" to a proper FQDN.
>> or
>> - Use module mod_rewrite to modify the 3xx error message returned by the server.
>>
>> So what I need to do in nginx to fix this issue? ?Thank you.
>
> If you proxy server via nginx, then probably something like this:
>
> server {
> ? ?server_name ?FQDN;
>
> ? ?location / {
> ? ? ? ?proxy_pass ? ? ?http://backend;
> ? ? ? ?proxy_redirect ?http://FQDN/ ?http://backend/;
> ? ?}
> }
>
>
> --
> Igor Sysoev
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>


From tomlove at gmail.com  Sat May 28 17:21:49 2011
From: tomlove at gmail.com (Thomas Love)
Date: Sat, 28 May 2011 15:21:49 +0200
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <405146017f0e366230d436dba381c983.NginxMailingListEnglish@forum.nginx.org>
References: <6f56cffc73aa96c01438b44bd1386150.NginxMailingListEnglish@forum.nginx.org>
	<405146017f0e366230d436dba381c983.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTinOvjSM7QRL+9RtPcu4LJORufFhGQ@mail.gmail.com>

On 27 May 2011 18:59, pk899 <nginx-forum at nginx.us> wrote:

> thanks. i have reading about nginx a lot on the internet. and it seems
> to me that many features of mod_security are available inside nginx.
>
> i like this post:
> http://eng.eelcowesemann.nl/linux-unix/nginx/nginx-blocking/
>
> my questions.
> 1. how should i block remote_addr based on IP in a certain external
> file?
>

It's in your link: include /etc/nginx/block.conf;
But it's not really nginx's job to secure your server from attackers. Use
iptables or similar for that.

2. to block xss or such type of injection attacks -- any best practice
> rules?
>

These are application-level problems (HTML, JS, SQL etc.) and are not
specific to your server software.


> 3. can i block based on "request_uri"? if some hosts are in it
> (parameter) then block it.
>

http://eng.eelcowesemann.nl/linux-unix/nginx/nginx-blocking/
Use $request_uri instead of $http_referer


> 4. can i block based on RBL checks? from spamhaus etc.


Not that I know of, and if you are deploying nginx for speed then this is
the last job you want to give it. Block proxies/spammers at the application
level where posting actually happens (e.g. in postcommentform.php), not for
every HTTP request you receive.


>

thanks!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110528/84ba52b1/attachment.html>

From tomlove at gmail.com  Sat May 28 17:29:57 2011
From: tomlove at gmail.com (Thomas Love)
Date: Sat, 28 May 2011 15:29:57 +0200
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <63174a38ac8500e2ac91fc3b3255a650.NginxMailingListEnglish@forum.nginx.org>
References: <6f56cffc73aa96c01438b44bd1386150.NginxMailingListEnglish@forum.nginx.org>
	<63174a38ac8500e2ac91fc3b3255a650.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTikDpDmvShNpuuPUoz0Pr31RfmcRZg@mail.gmail.com>

On 28 May 2011 05:54, pk899 <nginx-forum at nginx.us> wrote:

> Some problems with your recommendation, Igor.
>
> 1. There is no directive like "gzip_static".
>

http://wiki.nginx.org/NginxHttpGzipStaticModule


> 2. The proxy_pass that you recommend above -- this does a full 301
> redirect to Apache!


> This is really not the solution I am after.
>
> So, back to my first question, how do I pass the baton from Apache
> (which really needs to be my first front server because of all the
> features) only for the static files to nginx?
>

Then you set up mod_proxy on Apache instead. But with all due respect I
think you're after the wrong solution. You're unlikely to gain any
performance like that, because you're still making Apache do all the work of
talking to clients and holding keepalives. This is where nginx beats the
pants of Apache.

Put nginx in front. Test it by starting it on port 81 as suggested
previously. When you're happy, switch the ports (nginx on 80, proxying to
Apache on 127.0.0.1:81).


Thomas
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110528/8e003eb0/attachment.html>

From nginx-forum at nginx.us  Sat May 28 17:31:18 2011
From: nginx-forum at nginx.us (pk899)
Date: Sat, 28 May 2011 09:31:18 -0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <BANLkTinOvjSM7QRL+9RtPcu4LJORufFhGQ@mail.gmail.com>
References: <BANLkTinOvjSM7QRL+9RtPcu4LJORufFhGQ@mail.gmail.com>
Message-ID: <8adb1101582be7f16c5b6594f1b89dc3.NginxMailingListEnglish@forum.nginx.org>

Thanks. The RBL checking in Apache (via mod security) can happen in very
specific manner. I could specify that they check only specific arguments
(the very precise "input" field in the html) in a very specific page
("postcommentform.php"). 

Similarly, sure, the application needs to be smartly coded to prevent
against injections. But mod_security enables blocking this at a much
earlier phase in the web transaction. And it's easy to control this a
bit better at the hosting level. 

Clearly, I am looking at nginx not only as a "speed option", but as a
replacement for Apache. Several blogs online say that they have moved to
nginx. I am trying to see how. Apache is sadly bloated but thanks to
mod_security etc it's a very, very practical modern solution. 

Anyway, my setup above is not working either. Even just to use nginx as
merely a static server.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201449,201779#msg-201779



From tomlove at gmail.com  Sat May 28 17:34:46 2011
From: tomlove at gmail.com (Thomas Love)
Date: Sat, 28 May 2011 15:34:46 +0200
Subject: Web Server Internal IP Address/Internal Network Name Disclosure
	Vulnerability
In-Reply-To: <BANLkTinXKgKd9f3v7Sze=46y=QnmagxuWg@mail.gmail.com>
References: <BANLkTin0XJ=6GDoFJeJMQAO8VwJZUMs48A@mail.gmail.com>
	<20110527165054.GF46101@sysoev.ru>
	<BANLkTinXKgKd9f3v7Sze=46y=QnmagxuWg@mail.gmail.com>
Message-ID: <BANLkTikt7+rGHtO=npt-9asxsEe78AHtUA@mail.gmail.com>

On 28 May 2011 06:02, Mon Yang <treomon at gmail.com> wrote:

> Sorry...  FQDN = nginx server itself or the backend mail server?
>
> nginx server. The full name you would access the site with from the
outsite.


> Or, can you tell me how to change 3xx error message, since it might be
> safer so it won't break things?
>
> http://wiki.nginx.org/HttpCoreModule#error_page


> Thank you.
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110528/4bc98578/attachment.html>

From nginx-forum at nginx.us  Sat May 28 17:55:10 2011
From: nginx-forum at nginx.us (pk899)
Date: Sat, 28 May 2011 09:55:10 -0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <BANLkTikDpDmvShNpuuPUoz0Pr31RfmcRZg@mail.gmail.com>
References: <BANLkTikDpDmvShNpuuPUoz0Pr31RfmcRZg@mail.gmail.com>
Message-ID: <35e7eea4725030d90f826fb657beb2d3.NginxMailingListEnglish@forum.nginx.org>

Thanks. But I am a bit confused, as the simple setup is not working from
port 81 (for testing): 


  server {
    listen 81;
    server_name MYDOMAIN.com www.MYDOMAIN.com;

    location ^~ /site/static {
      root /home/MYDOMAIN/www/static;
      index index.htm index.php;
      expires       30d;
      gzip          on;
    }
    
    location / {
      proxy_pass      http://127.0.0.1:80/;
      proxy_redirect  off;
      
      proxy_set_header   Host             $host;
      proxy_set_header   X-Forwarded-For  $proxy_add_x_forwarded_for;
    }

  }



Then I try in my browser this: 
http://MYDOMAIN.com:81/test.gif    

On the server the "test.gif" file is at
"/home/MYDOMAIN/www/static/test.gif". 

Questions: 

(1) What am I missing? Should the test.gif file be somewhere else?
Should I try this with some other URL instead of the one I mention
above? 

(2) Also, if this worked, and I accepted Nginx to be at the front and
moved Apache to the background, isn't it true that Apache would have to
pass back all processed output to Nginx? 

(3) Even with "proxy_redirect off", when I try http://MYDOMAIN.com:81,
it actually does a hard 301 redirect to my Apache server. 

What am I missing?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201449,201787#msg-201787



From tomlove at gmail.com  Sat May 28 17:56:41 2011
From: tomlove at gmail.com (Thomas Love)
Date: Sat, 28 May 2011 15:56:41 +0200
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <8adb1101582be7f16c5b6594f1b89dc3.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTinOvjSM7QRL+9RtPcu4LJORufFhGQ@mail.gmail.com>
	<8adb1101582be7f16c5b6594f1b89dc3.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTiku48XZ=NPaWP16HdwmENTLBebPXA@mail.gmail.com>

On 28 May 2011 15:31, pk899 <nginx-forum at nginx.us> wrote:

> Thanks. The RBL checking in Apache (via mod security) can happen in very
> specific manner. I could specify that they check only specific arguments
> (the very precise "input" field in the html) in a very specific page
> ("postcommentform.php").
>

There is no module like this for nginx that I'm aware of but you can
probably find application-level libraries to do it. The fewer conditionals
at the HTTP server level the better.


> Similarly, sure, the application needs to be smartly coded to prevent
> against injections. But mod_security enables blocking this at a much
> earlier phase in the web transaction. And it's easy to control this a
> bit better at the hosting level.
>
> If it's important then leave it on Apache with your web application. It
will still do what it does upstream of nginx.


> Clearly, I am looking at nginx not only as a "speed option", but as a
> replacement for Apache. Several blogs online say that they have moved to
> nginx. I am trying to see how. Apache is sadly bloated but thanks to
> mod_security etc it's a very, very practical modern solution.
>

IMO its bloat is not helped by being piled with high with modules. This is
where nginx comes in, and being relatively free of modules was one of its
selling points for me.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110528/03bec3e2/attachment.html>

From tomlove at gmail.com  Sat May 28 18:05:38 2011
From: tomlove at gmail.com (Thomas Love)
Date: Sat, 28 May 2011 16:05:38 +0200
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <35e7eea4725030d90f826fb657beb2d3.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTikDpDmvShNpuuPUoz0Pr31RfmcRZg@mail.gmail.com>
	<35e7eea4725030d90f826fb657beb2d3.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <BANLkTi=PqSinGCphPxEnq-m11RpvS+Y6mA@mail.gmail.com>

On 28 May 2011 15:55, pk899 <nginx-forum at nginx.us> wrote:

> Thanks. But I am a bit confused, as the simple setup is not working from
> port 81 (for testing):
>
>
>  server {
>    listen 81;
>    server_name MYDOMAIN.com www.MYDOMAIN.com;
>
>    location ^~ /site/static {
>      root /home/MYDOMAIN/www/static;
>      index index.htm index.php;
>      expires       30d;
>      gzip          on;
>    }
>
>    location / {
>      proxy_pass      http://127.0.0.1:80/;
>      proxy_redirect  off;
>
>      proxy_set_header   Host             $host;
>      proxy_set_header   X-Forwarded-For  $proxy_add_x_forwarded_for;
>    }
>
>  }
>
> server {
   listen  81;
   server_name MYDOMAIN.com www.MYDOMAIN.com <http://www.mydomain.com/>;

   root  /home/MYDOMAIN/www;

   location / {
       proxy_pass        http://127.0.0.1:80;
       proxy_set_header  Host  $host;
       proxy_set_header  X-Forwarded-For  $proxy_add_x_forwarded_for;
   }

   location /static/ {
       index index.htm;
       expires      30d;
       gzip         on;
       gzip_static  on;
   }

}
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110528/bb330d26/attachment-0001.html>

From nginx-forum at nginx.us  Sat May 28 18:18:51 2011
From: nginx-forum at nginx.us (pk899)
Date: Sat, 28 May 2011 10:18:51 -0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <BANLkTi=PqSinGCphPxEnq-m11RpvS+Y6mA@mail.gmail.com>
References: <BANLkTi=PqSinGCphPxEnq-m11RpvS+Y6mA@mail.gmail.com>
Message-ID: <1ef32b04b3a6e62585463e4f71c6aa1e.NginxMailingListEnglish@forum.nginx.org>

Thanks. 

So with this setup, this works: 

http://mydomain.com:81/static/test.gif 

But when I try this: 

http://mydomain.com:81/static/test.gif 

It physically redirects to http://mydomain.com. 

How can I avoid this?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201449,201792#msg-201792



From nginx-forum at nginx.us  Sat May 28 18:20:30 2011
From: nginx-forum at nginx.us (pk899)
Date: Sat, 28 May 2011 10:20:30 -0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <1ef32b04b3a6e62585463e4f71c6aa1e.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTi=PqSinGCphPxEnq-m11RpvS+Y6mA@mail.gmail.com>
	<1ef32b04b3a6e62585463e4f71c6aa1e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <de07e2d0f619854ff098aa8181cec39e.NginxMailingListEnglish@forum.nginx.org>

Because this forum does not allow editing, here's my edit: 

Thanks.

So with this setup, this works:

http://mydomain.com:81/static/test.gif

But when I try this:

http://mydomain.com:81/     (main site) 

It **physically** redirects to http://mydomain.com.

How can I avoid this?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201449,201793#msg-201793



From nginx-forum at nginx.us  Sat May 28 18:28:52 2011
From: nginx-forum at nginx.us (pk899)
Date: Sat, 28 May 2011 10:28:52 -0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <BANLkTinOvjSM7QRL+9RtPcu4LJORufFhGQ@mail.gmail.com>
References: <BANLkTinOvjSM7QRL+9RtPcu4LJORufFhGQ@mail.gmail.com>
Message-ID: <3b8030456b07dfcd74fb2bc67fde0138.NginxMailingListEnglish@forum.nginx.org>

Thomas Love Wrote:
-------------------------------------------------------


> > 1. how should i block remote_addr based on IP in
> a certain external
> > file?
> >
> 
> It's in your link: include /etc/nginx/block.conf;


Not correct. The "block.conf" as that guy mentions is only a collection
of rules. 

I am asking if I can have a file that is full of host names or IPs, say
"blacklist.conf". Then, I want to do something like this: 


if ($request_uri ~* ("/blacklist.conf") ) {
  return 403;
}

if ($remote_addr ~* ("/blacklist.conf") ) {
  return 403;
}


This way I can keep easily adding or removing IPs/hosts from my
blacklist file instead of editing the nginx.conf file every time. 

Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201449,201796#msg-201796



From mdounin at mdounin.ru  Sat May 28 18:32:35 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sat, 28 May 2011 18:32:35 +0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <35e7eea4725030d90f826fb657beb2d3.NginxMailingListEnglish@forum.nginx.org>
References: <BANLkTikDpDmvShNpuuPUoz0Pr31RfmcRZg@mail.gmail.com>
	<35e7eea4725030d90f826fb657beb2d3.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110528143235.GJ42265@mdounin.ru>

Hello!

On Sat, May 28, 2011 at 09:55:10AM -0400, pk899 wrote:

> Thanks. But I am a bit confused, as the simple setup is not working from
> port 81 (for testing): 
> 
> 
>   server {
>     listen 81;
>     server_name MYDOMAIN.com www.MYDOMAIN.com;
> 
>     location ^~ /site/static {
>       root /home/MYDOMAIN/www/static;
>       index index.htm index.php;
>       expires       30d;
>       gzip          on;
>     }
>     
>     location / {
>       proxy_pass      http://127.0.0.1:80/;
>       proxy_redirect  off;
>       
>       proxy_set_header   Host             $host;
>       proxy_set_header   X-Forwarded-For  $proxy_add_x_forwarded_for;
>     }
> 
>   }
> 
> 
> 
> Then I try in my browser this: 
> http://MYDOMAIN.com:81/test.gif    
> 
> On the server the "test.gif" file is at
> "/home/MYDOMAIN/www/static/test.gif". 
> 
> Questions: 
> 
> (1) What am I missing? Should the test.gif file be somewhere else?
> Should I try this with some other URL instead of the one I mention
> above? 

It looks like you don't understand at least two basic things:

1. Location directive maps *URI* namespace to configuration.  
That is, with your config /test.gif will be mapped to "location 
/", i.e. will be proxy_pass'ed to Apache.  You want to request 
something like /site/static/test.gif to be actually served by 
nginx itself.

See here for more details:

http://wiki.nginx.org/HttpCoreModule#location

2. Root directive define *root*, and URI will be added to it to 
map request to filesystem.  I.e. with URI /site/static/test.gif 
and root /home/MYDOMAIN/www/static you will end up with 
"/home/MYDOMAIN/www/static/site/static/test.gif" filename.

If you want "/site/static/test.gif" to be mapped to 
"/home/MYDOMAIN/www/static/test.gif" you have to use alias 
directive instead.

See here for more details:

http://wiki.nginx.org/HttpCoreModule#root
http://wiki.nginx.org/HttpCoreModule#alias

> (2) Also, if this worked, and I accepted Nginx to be at the front and
> moved Apache to the background, isn't it true that Apache would have to
> pass back all processed output to Nginx? 

Yes.  And this is a huge win in many cases: Apache processes won't 
be bound by serving responses to slow clients.

> (3) Even with "proxy_redirect off", when I try http://MYDOMAIN.com:81,
> it actually does a hard 301 redirect to my Apache server. 

Directive proxy_redirect is to fix redirects returned by backend 
server.  By switching it off you merely said "don't touch 
anything" to nginx, and this is what it does.  I.e. redirect is 
returned by your Apache server (alternatively, it may be just your 
browser's cache from previous testing - check logs).

Maxim Dounin


From nginx-forum at nginx.us  Sun May 29 09:11:09 2011
From: nginx-forum at nginx.us (pk899)
Date: Sun, 29 May 2011 01:11:09 -0400
Subject: Newbie questions about nginx (moving from apache)
In-Reply-To: <6f56cffc73aa96c01438b44bd1386150.NginxMailingListEnglish@forum.nginx.org>
References: <6f56cffc73aa96c01438b44bd1386150.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <70a981027302704e080d171c450740c4.NginxMailingListEnglish@forum.nginx.org>

thanks igor and maxim. the proxy stuff is not working for me but i am so
liking the nginx server that i will follow your advice to not care about
mod security so much and just do application level checking in php. 

my question: is there a simple guide to get php working on nginx when it
is already installed and running with apache as mod_php on the server?
while testing nginx i dont want to break already working functionality.


there are online guides for "fastcgi" but they are all different and
sometimes old syntax so they are confusing. is there a recommended
official config file that will cover the best practice for a production
server, including: 

1. static files be served with "expires" or etags etc
2. php with all the modules installed on my server (memcache, etc) 
3. secure php 

i want to now just try nginx running on server:81 port and see if it can
do *everything* that apache does, if i move some of the functionality to
my php application (such as checking RBLs etc).

thanks so much.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201449,201911#msg-201911



From nginx-forum at nginx.us  Sun May 29 18:29:07 2011
From: nginx-forum at nginx.us (pk899)
Date: Sun, 29 May 2011 10:29:07 -0400
Subject: Error message control when using nginx as proxy..
Message-ID: <d0fad90718b4c4e27292d5cc957fcf44.NginxMailingListEnglish@forum.nginx.org>

I have setup nginx as a front proxy, it passes php requests to Apache.

When I now try to access my site through, say, "TRACE", then naturally I
see the 405 error. But this is a nginx default message. How can I make
sure that I see my own customized error page? 

I have this in my "Server" definition...

  #================================
  # Custom error page...
  #================================
  error_page 400 401 402 403 405 406 407 408 409 410 411 412 413 414 415
416 417 495 496 497
             500 501 502 503 504 505 506 507 /error.htm;

But this does not seem to be working. For the 405, it still shows the
default nginx page. Anything I can do about this? 

Secondly, for Apache, there is still some mod_security stuff that sends
back a "406". How can I make sure that Apache sending that signal back
to Nginx only triggers the 406 I have setup in nginx? 

Thanks!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,201978,201978#msg-201978



From nginx-forum at nginx.us  Sun May 29 20:07:56 2011
From: nginx-forum at nginx.us (pk899)
Date: Sun, 29 May 2011 12:07:56 -0400
Subject: Limit_conn with zones -- safe production server value?
Message-ID: <81e2c45c150a8e0ac12ed3dd9f1b1503.NginxMailingListEnglish@forum.nginx.org>

Hi. I am trying to replicate the functionality of mod_evasive on Apache,
which basically says that at the same point the same IP can have only "X
connections per second". 

I see the example here: 
http://wiki.nginx.org/HttpLimitZoneModule

But this uses "binary_remote_addr". How does this cater for shared IPs,
where people in the same network may actually have the same remote addr.


Can I therefore use some other variable, such as a combination of
"binary_remote_addr" and their "http_user_agent"? 

What is the most often used value in production environments? 

Thanks!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,202012,202012#msg-202012



From jackdesert556 at gmail.com  Mon May 30 01:03:19 2011
From: jackdesert556 at gmail.com (Jack Desert)
Date: Sun, 29 May 2011 17:03:19 -0400
Subject: Getting All Documentation Under One Roof
In-Reply-To: <4DE02376.4040503@gmail.com>
References: <20110527153818.2d53fda3@dime>
	<1306532334.31643.6.camel@portable-evil>
	<4DE02376.4040503@gmail.com>
Message-ID: <20110529170319.459f7d55@dime>

On Sat, 28 May 2011 01:19:34 +0300
Eugaia <ngx.eugaia at gmail.com> wrote:
 
> I agree that linking to the original docs would also be useful.


That's a good idea. I put a link to 
  http://nginx.org/en/docs/
on the 
  http://wiki.nginx.org/Resources
page

I've also been looking at the organization of the main docs on nginx.org. It appears that there are a total of seven documents that comprise the "Official Documentation", they being:

-How nginx processes a request
-Server names
-Configuring HTTPS servers
-(How to) debugging log
-(How to) converting rewrite rules
-nginx/windows usage
-(FAQ) sys_errlist message

However, it is difficult to see all these articles from a single web page. I don't know who manages the site, and noone has asked my opinion, but I believe it would be more intuitive to find nginx documentation if http://nginx.org/en/docs/ were a table of contents, laid out like this:

documentation
    introduction to nginx
        How nginx processes a request
        Server names
    How to
        Configuring HTTPS servers
        (Create a) debugging log
        converting rewrite rules
    security advisories
    nginx/windows usage
    FAQ

This way one could see all the options from a single, hierarchal glance. 

Since "security advisories", "howto", and "introduction" are all contained within this documentation, their links could be removed from the main page like this:

{languages}
news
about
download
documentation
faq
wiki
links
books
support

which would make the main page cleaner. 

Now that's a good start to making things easy to find. But I notice that if you look closely, there are additional documents floating around. For example, go to http://nginx.org/ru/ (The Russian "about" page), and the text under "Basic HTTP Features" (???????? ???????????????? HTTP-???????) contains not just plain text, but links to more documentation located at http://sysoev.ru/nginx/docs/. Upon reading some of these docs, it looks like they comprised the beginning of the (possibly original?) wiki. And since they've been put into the wiki and not into the official docs page, I wonder if perhaps the plan has been for the wiki to replace the master docs? What is the overall plan here for documentation? Which is meant to be the master docs, the wiki or ngingx.org/docs? 

I want to help the cause by getting all the documentation under one roof, easily accessible, and well-organized. Help me to understand how I can do this. 

Jack Desert


From tomlove at gmail.com  Mon May 30 02:44:40 2011
From: tomlove at gmail.com (Thomas Love)
Date: Mon, 30 May 2011 00:44:40 +0200
Subject: Getting All Documentation Under One Roof
In-Reply-To: <20110529170319.459f7d55@dime>
References: <20110527153818.2d53fda3@dime>
	<1306532334.31643.6.camel@portable-evil>
	<4DE02376.4040503@gmail.com> <20110529170319.459f7d55@dime>
Message-ID: <BANLkTimy2MrXK2nKzdfb3WxJAu_SfUyTrA@mail.gmail.com>

On 29 May 2011 23:03, Jack Desert <jackdesert556 at gmail.com> wrote:

> On Sat, 28 May 2011 01:19:34 +0300
> Eugaia <ngx.eugaia at gmail.com> wrote:
>
>
> I want to help the cause by getting all the documentation under one roof,
> easily accessible, and well-organized. Help me to understand how I can do
> this.
>
> Jack Desert
>
>
I'd be in favour of copying the pages of http://nginx.org/en/docs/ into
http://wiki.nginx.org/ along with links to security advisories, books, etc.
It would need to be kept up-to-date, but this would be a start, unless Igor
is keen to keep them separate.

I suspect there may be a better way to organise the top-level wiki sections
(Install, Modules, Add-ons etc.) e.g. I was surprised to find
http://wiki.nginx.org/HttpCoreModule in Modules when I first visited (it's
logical, but wasn't obvious). I also think "Install" on the top level can be
reworked to incorporate "Getting started" and "Full example", modified to
implement auth_basic as well since this can easily be done dangerously by
first-timers.

So I would love to see two navigation paths streamlined: the path first-time
users take in downloading, setting up, configuring, and the path returning
users take getting reference on core, or on modules. Would be happy to help
implement, if this makes sense to anyone else.

Thomas
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110530/b6d9bd0e/attachment.html>

From Dalibor.Jotanovic at herold.at  Mon May 30 11:02:14 2011
From: Dalibor.Jotanovic at herold.at (Dalibor Jotanovic)
Date: Mon, 30 May 2011 09:02:14 +0200
Subject: Antw: Re: NGINX loadbalancing ORACLE forms backends?
In-Reply-To: <BANLkTi=us=EwWDtsKehK-LVQH7LkXkWH5w@mail.gmail.com>
References: <4DDF9C640200005E0003480F@mx.herold.at>
	<BANLkTi=us=EwWDtsKehK-LVQH7LkXkWH5w@mail.gmail.com>
Message-ID: <4DE35D16.97C0.005E.0@herold.at>

Hi,

i'll take a look at it. Thanks for the hint.

Cheers,
DJ

>>> J?r?me Loyet<jerome at loyet.net> 5/27/2011 12:52 >>>
Hi


have a look on this module I've made: http://code.google.com/p/nginx-sticky-module/


I'm using it in different environments (PHP, tomcat, jboss, ...). It can feet your need. Moreover it's independant from the type of backend you're using.


hope this help


++ Jerome


2011/5/27 Dalibor Jotanovic <Dalibor.Jotanovic at herold.at>

Hi,
does anybody have a working setup with NGINX loadbalancing ORACLE FORMS backends?
I'm interested in setting up an environment with JSESSION cookie based sticky sessions.
Regards,
DJ





HEROLD Business Data GmbH
Guntramsdorfer Stra?e 105
A-2340 M?dling

FN 233171z
Landesgericht Wiener Neustadt

Besuchen Sie uns online und mobil auf www.herold.at!

Weitere Informationen zu unseren Produkten finden Sie unter:
http://mobile.herold.at
http://daten.herold.at 

Werden Sie Fan von HEROLD auf Facebook!


Bitte beachten Sie auch unsere Tochter- und Partner-Unternehmen:
www.clearsense.at 
www.tupalo.com 
www.aboutmedia.at 
www.love.at 
www.reise-hero.at 
www.urlauburlaub.at 

Diese E-Mail kann vertrauliche und/oder rechtlich gesch?tzte Informationen enthalten. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrt?mlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie die unbefugte Weitergabe dieser Mail sind nicht gestattet.

This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and delete this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden. 

_______________________________________________
nginx mailing list
nginx at nginx.org
http://nginx.org/mailman/listinfo/nginx







Click here to report this email as spam.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110530/b44370d7/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1993 bytes
Desc: not available
URL: <http://nginx.org/pipermail/nginx/attachments/20110530/b44370d7/attachment-0002.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1096 bytes
Desc: not available
URL: <http://nginx.org/pipermail/nginx/attachments/20110530/b44370d7/attachment-0003.gif>

From Dalibor.Jotanovic at herold.at  Mon May 30 11:04:02 2011
From: Dalibor.Jotanovic at herold.at (Dalibor Jotanovic)
Date: Mon, 30 May 2011 09:04:02 +0200
Subject: Antw: Re: NGINX loadbalancing ORACLE forms backends?
In-Reply-To: <BANLkTi=m1zbSLoP=3S-Q0gPxUbBJ39_P+w@mail.gmail.com>
References: <4DDF9C640200005E0003480F@mx.herold.at>
	<BANLkTi=m1zbSLoP=3S-Q0gPxUbBJ39_P+w@mail.gmail.com>
Message-ID: <4DE35D820200005E000348CF@mx.herold.at>

Hi,
 
i am actually using this module at the moment, but the documentated
config option "srun...."  doesn't seem to work in the latest nginx
version. Does the option work on your side? Don't you get any errors?
 
Cheers,
DJ

>>> ???<nbubingo at gmail.com> 5/27/2011 1:20 >>>
See also this module:
http://code.google.com/p/nginx-upstream-jvm-route/

 ( http://code.google.com/p/nginx-upstream-jvm-route/ )This module
achieves session stickiness with the session cookie. If the session is
not in the cookie or URL, the module will be a normal Round-Robin
upstream module.

2011/5/27 Dalibor Jotanovic <Dalibor.Jotanovic at herold.at>


Hi,
does anybody have a working setup with NGINX loadbalancing ORACLE FORMS
backends?
I'm interested in setting up an environment with JSESSION cookie based
sticky sessions.
Regards,
DJ

 ( http://www.herold.at )
 ( http://www.herold.at )

HEROLD Business Data GmbH
Guntramsdorfer Stra?e 105
A-2340 M?dling

FN 233171z
Landesgericht Wiener Neustadt

Besuchen Sie uns online und mobil auf www.herold.at!

Weitere Informationen zu unseren Produkten finden Sie unter:
http://mobile.herold.at
http://daten.herold.at 

 ( http://www.facebook.com/derherold )Werden Sie Fan von HEROLD auf
Facebook!


Bitte beachten Sie auch unsere Tochter- und Partner-Unternehmen:
www.clearsense.at 
www.tupalo.com 
www.aboutmedia.at  ( http://www.tupalo.com )
www.love.at  ( http://www.tupalo.com )
www.reise-hero.at  ( http://www.tupalo.com )
www.urlauburlaub.at  ( http://www.tupalo.com )
 ( http://www.tupalo.com )
Diese E-Mail kann vertrauliche und/oder rechtlich gesch?tzte
Informationen enthalten. Wenn Sie nicht der richtige Adressat sind oder
diese E-Mail irrt?mlich erhalten haben, informieren Sie bitte sofort den
Absender und vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie
die unbefugte Weitergabe dieser Mail sind nicht gestattet.

This e-mail may contain confidential and/or privileged information. If
you are not the intended recipient (or have received this e-mail in
error) please notify the sender immediately and delete this e-mail. Any
unauthorized copying, disclosure or distribution of the material in this
e-mail is strictly forbidden. 

_______________________________________________
nginx mailing list
nginx at nginx.org
http://nginx.org/mailman/listinfo/nginx






Click here (
https://www.mailcontrol.com/sr/t91yjfYfe5jTndxI!oX7Uh5Cv+dgaYj8ylv+aQFyzJM2yPGdnNdCJu57SCsCe4JD1GQ2ktvKI6YpdY6WjzZSeg==
) to report this email as spam.
HEROLD Business Data GmbH
Guntramsdorfer Stra?e 105
A-2340 M?dling

FN 233171z
Landesgericht Wiener Neustadt 
  
Besuchen Sie uns online und mobil www.herold.at!

Weitere Informationen zu unseren Produkten finden Sie unter:
http://mobile.herold.at
http://daten.herold.at

Werden Sie Fan von HEROLD auf Facebook!


Bitte beachten Sie auch unsere Tochter- und Partner-Unternehmen:
www.clearsense.at
www.tupalo.com
www.aboutmedia.at
www.love.at
www.reise-hero.at
www.urlauburlaub.at
 
Diese E-Mail kann vertrauliche und/oder rechtlich gesch?tzte
Informationen enthalten. Wenn Sie nicht der richtige Adressat sind oder
diese E-Mail irrt?mlich erhalten haben, informieren Sie bitte sofort den
Absender und vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie
die unbefugte Weitergabe dieser Mail sind nicht gestattet. 

This e-mail may contain confidential and/or privileged information. If
you are not the intended recipient (or have received this e-mail in
error) please notify the sender immediately and delete this e-mail. Any
unauthorized copying, disclosure or distribution of the material in this
e-mail is strictly forbidden. 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110530/5f038603/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1993 bytes
Desc: CompuServe GIF graphic
URL: <http://nginx.org/pipermail/nginx/attachments/20110530/5f038603/attachment.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1096 bytes
Desc: CompuServe GIF graphic
URL: <http://nginx.org/pipermail/nginx/attachments/20110530/5f038603/attachment-0001.gif>

From davidkazuhiro at gmail.com  Mon May 30 11:32:19 2011
From: davidkazuhiro at gmail.com (David Somers-Harris)
Date: Mon, 30 May 2011 16:32:19 +0900
Subject: HttpImageFilterModule: Crop with nonzero origin
Message-ID: <BANLkTik5zeVtGp4kH=WrEfFVrA5pAjsWgQ@mail.gmail.com>

I'd like to use nginx's Image Filter
Module<http://wiki.nginx.org/HttpImageFilterModule>to dynamically
generate
DeepZoom<http://msdn.microsoft.com/en-us/library/cc645050%28v=vs.95%29.aspx>images.

However, I don't see an x-position or y-position argument for the crop
directive. Is it even possible for the Image Filter Module to crop quadrants
other than the top left?

David
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110530/c59afa8d/attachment.html>

From Dalibor.Jotanovic at herold.at  Mon May 30 13:27:11 2011
From: Dalibor.Jotanovic at herold.at (Dalibor Jotanovic)
Date: Mon, 30 May 2011 11:27:11 +0200
Subject: Antw: Re: NGINX loadbalancing ORACLE forms backends?
In-Reply-To: <4DE35D820200005E000348CF@mx.herold.at>
References: <4DDF9C640200005E0003480F@mx.herold.at>
	<BANLkTi=m1zbSLoP=3S-Q0gPxUbBJ39_P+w@mail.gmail.com>
	<4DE35D820200005E000348CF@mx.herold.at>
Message-ID: <4DE37F0F.97C0.005E.0@herold.at>

Hi again,

iirc that jvm route plugin is not supported by nginx. Anyway if somebody is having the same issue i had with that plugin and the mentioned srun config option, here is the *possible* solution to it:

Set the jvm_route $cookie JESSIONID;  before the server and srun definitons. Otherwise the plugin is not loaded and the srun option is unknown i guess...

Cheers,
DJ


>>> "Dalibor Jotanovic" <Dalibor.Jotanovic at herold.at> 5/30/2011 9:04 >>>

Hi,

i am actually using this module at the moment, but the documentated config option "srun...."  doesn't seem to work in the latest nginx version. Does the option work on your side? Don't you get any errors?

Cheers,
DJ

>>> ???<nbubingo at gmail.com> 5/27/2011 1:20 >>>
See also this module: http://code.google.com/p/nginx-upstream-jvm-route/


This module achieves session stickiness with the session cookie. If the session is not in the cookie or URL, the module will be a normal Round-Robin upstream module.



2011/5/27 Dalibor Jotanovic <Dalibor.Jotanovic at herold.at>

Hi,
does anybody have a working setup with NGINX loadbalancing ORACLE FORMS backends?
I'm interested in setting up an environment with JSESSION cookie based sticky sessions.
Regards,
DJ





HEROLD Business Data GmbH
Guntramsdorfer Stra?e 105
A-2340 M?dling

FN 233171z
Landesgericht Wiener Neustadt

Besuchen Sie uns online und mobil auf www.herold.at!

Weitere Informationen zu unseren Produkten finden Sie unter:
http://mobile.herold.at
http://daten.herold.at 

Werden Sie Fan von HEROLD auf Facebook!


Bitte beachten Sie auch unsere Tochter- und Partner-Unternehmen:
www.clearsense.at 
www.tupalo.com 
www.aboutmedia.at 
www.love.at 
www.reise-hero.at 
www.urlauburlaub.at 

Diese E-Mail kann vertrauliche und/oder rechtlich gesch?tzte Informationen enthalten. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrt?mlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie die unbefugte Weitergabe dieser Mail sind nicht gestattet.

This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and delete this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden. 

_______________________________________________
nginx mailing list
nginx at nginx.org
http://nginx.org/mailman/listinfo/nginx







Click here to report this email as spam.





HEROLD Business Data GmbH
Guntramsdorfer Stra?e 105
A-2340 M?dling

FN 233171z
Landesgericht Wiener Neustadt

Besuchen Sie uns online und mobil auf www.herold.at!

Weitere Informationen zu unseren Produkten finden Sie unter:
http://mobile.herold.at
http://daten.herold.at 

Werden Sie Fan von HEROLD auf Facebook!


Bitte beachten Sie auch unsere Tochter- und Partner-Unternehmen:
www.clearsense.at 
www.tupalo.com 
www.aboutmedia.at 
www.love.at 
www.reise-hero.at 
www.urlauburlaub.at 

Diese E-Mail kann vertrauliche und/oder rechtlich gesch?tzte Informationen enthalten. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrt?mlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie die unbefugte Weitergabe dieser Mail sind nicht gestattet.

This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and delete this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden. 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110530/64837402/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1993 bytes
Desc: not available
URL: <http://nginx.org/pipermail/nginx/attachments/20110530/64837402/attachment.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1096 bytes
Desc: not available
URL: <http://nginx.org/pipermail/nginx/attachments/20110530/64837402/attachment-0001.gif>

From mdounin at mdounin.ru  Mon May 30 13:50:04 2011
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 30 May 2011 13:50:04 +0400
Subject: HttpImageFilterModule: Crop with nonzero origin
In-Reply-To: <BANLkTik5zeVtGp4kH=WrEfFVrA5pAjsWgQ@mail.gmail.com>
References: <BANLkTik5zeVtGp4kH=WrEfFVrA5pAjsWgQ@mail.gmail.com>
Message-ID: <20110530095003.GL42265@mdounin.ru>

Hello!

On Mon, May 30, 2011 at 04:32:19PM +0900, David Somers-Harris wrote:

> I'd like to use nginx's Image Filter
> Module<http://wiki.nginx.org/HttpImageFilterModule>to dynamically
> generate
> DeepZoom<http://msdn.microsoft.com/en-us/library/cc645050%28v=vs.95%29.aspx>images.
> 
> However, I don't see an x-position or y-position argument for the crop
> directive. Is it even possible for the Image Filter Module to crop quadrants
> other than the top left?

Just a side note: "crop" isn't really crop, it is something like 
"resize and crop if needed".

And no, it doesn't allow to crop image to arbitrary selected region.

Maxim Dounin


From nginx-forum at nginx.us  Mon May 30 17:40:09 2011
From: nginx-forum at nginx.us (atipico)
Date: Mon, 30 May 2011 09:40:09 -0400
Subject: =?UTF-8?Q?Simple_URI_redirection_/_alias_in_Nginx=E2=80=A6?=
Message-ID: <841ddad79482575f6eecec727b5e01de.NginxMailingListEnglish@forum.nginx.org>

Please, how can I make this simple operation:

I wish that every requests for uris like: 
	http://example.com/adm
or:
	http://example.com/adm
Goes to the wordpress login page:
	wordpress/wp-login.php
(The wordpress installation is inside the folder "wordpress").

I've tried:

	location /adm/ { alias wordpress/wp-login.php; }
	location /adm { alias wordpress/wp-login.php; }

and:

	rewrite ^/adm$ wordpress/wp-login.php;

But with no success...

Just in case someone asks, this is my serve conf. file:

	server {
		server_name www.example.com;
		rewrite ^ http://example.com$request_uri? permanent;
	}
	
	server {
		server_name example.com;
	
		access_log /var/log/nginx/example.com.access;
		error_log /var/log/nginx/example.com.error;
	
		root /var/www/example.com;
		
		index index.htm index.php;
	
		location / {try_files $uri /wp$uri/ /wordpress/index.php$args;}
	
		location ~ .php$ {
			try_files $uri =404;
			#fastcgi_split_path_info ^(.+\.php)(/.+)$;
			fastcgi_pass 127.0.0.1:9000;
			fastcgi_index index.php;
			fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
			include /etc/nginx/fastcgi_params;
		}
			
		location = /favicon.ico {log_not_found off;access_log off;}
		location = /robots.txt {allow all;log_not_found off;access_log off;}
		location ~ /\. {deny all;access_log off;log_not_found off;	}
	}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,202235,202235#msg-202235



From igor at sysoev.ru  Mon May 30 22:47:10 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Mon, 30 May 2011 22:47:10 +0400
Subject: =?UTF-8?Q?Re=3A_Simple_URI_redirection_/_alias_in_Nginx=E2=80=A6?=
In-Reply-To: <841ddad79482575f6eecec727b5e01de.NginxMailingListEnglish@forum.nginx.org>
References: <841ddad79482575f6eecec727b5e01de.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110530184709.GD87429@sysoev.ru>

On Mon, May 30, 2011 at 09:40:09AM -0400, atipico wrote:
> Please, how can I make this simple operation:
> 
> I wish that every requests for uris like: 
> 	http://example.com/adm
> or:
> 	http://example.com/adm
> Goes to the wordpress login page:
> 	wordpress/wp-login.php
> (The wordpress installation is inside the folder "wordpress").
> 
> I've tried:
> 
> 	location /adm/ { alias wordpress/wp-login.php; }
> 	location /adm { alias wordpress/wp-login.php; }
> 
> and:
> 
> 	rewrite ^/adm$ wordpress/wp-login.php;
> 
> But with no success...
> 
> Just in case someone asks, this is my serve conf. file:
> 
> 	server {
> 		server_name www.example.com;
> 		rewrite ^ http://example.com$request_uri? permanent;
> 	}
> 	
> 	server {
> 		server_name example.com;
> 	
> 		access_log /var/log/nginx/example.com.access;
> 		error_log /var/log/nginx/example.com.error;
> 	
> 		root /var/www/example.com;
> 		
> 		index index.htm index.php;
> 	
> 		location / {try_files $uri /wp$uri/ /wordpress/index.php$args;}
> 	
> 		location ~ .php$ {
> 			try_files $uri =404;
> 			#fastcgi_split_path_info ^(.+\.php)(/.+)$;
> 			fastcgi_pass 127.0.0.1:9000;
> 			fastcgi_index index.php;
> 			fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
> 			include /etc/nginx/fastcgi_params;
> 		}
> 			
> 		location = /favicon.ico {log_not_found off;access_log off;}
> 		location = /robots.txt {allow all;log_not_found off;access_log off;}
> 		location ~ /\. {deny all;access_log off;log_not_found off;	}
> 	}

location = /adm/ {
    fastcgi_pass 127.0.0.1:9000;
    fastcgi_param SCRIPT_FILENAME /var/www/example.com/wordpress/wp-login.php
    include /etc/nginx/fastcgi_params;
}

location = /adm {
    fastcgi_pass 127.0.0.1:9000;
    fastcgi_param SCRIPT_FILENAME /var/www/example.com/wordpress/wp-login.php
    include /etc/nginx/fastcgi_params;
}


-- 
Igor Sysoev


From fxn at hashref.com  Mon May 30 23:33:45 2011
From: fxn at hashref.com (Xavier Noria)
Date: Mon, 30 May 2011 21:33:45 +0200
Subject: mod_proxy, compression, and chunked responses
Message-ID: <BANLkTi=__rWitoVxNoTDW8sB9=VYouVyWg@mail.gmail.com>

I have an application server behind nginx that may output chunked
responses, and would like to use compression for them as I do for
ordinary responses.

Is that possible with nginx? If it is, is it possible in a way that
nginx compresses stuff on the fly and sends chunked (compressed)
responses itself rather than buffering the original response and
sending a non-chunked response itself? (That would defeat the purpose
of the chunked responses.)


From nginx-forum at nginx.us  Tue May 31 02:56:39 2011
From: nginx-forum at nginx.us (pk899)
Date: Mon, 30 May 2011 18:56:39 -0400
Subject: =?UTF-8?Q?Re=3A_Simple_URI_redirection_/_alias_in_Nginx=E2=80=A6?=
In-Reply-To: <841ddad79482575f6eecec727b5e01de.NginxMailingListEnglish@forum.nginx.org>
References: <841ddad79482575f6eecec727b5e01de.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <70b8525468895548826d5775ed0bf6f6.NginxMailingListEnglish@forum.nginx.org>

I have a similar issue. 

Do we need separate location directives for each and every possible
path, or can we do some "rewrite" inside location for the main / root
folder? For example: 

location = / {

  # rewrite rules come here..
  rewrite "^/adm$" "/wordpress/wp-login.php";
  ....
  fastcgi_pass 127.0.0.1:9000;
  fastcgi_param SCRIPT_FILENAME
/var/www/example.com/wordpress/wp-login.php
  include /etc/nginx/fastcgi_params;

}



To me this is cleaner, but want to make sure that this is not the less
desirable way? 

Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,202235,202430#msg-202430



From nginx-forum at nginx.us  Tue May 31 02:57:50 2011
From: nginx-forum at nginx.us (pk899)
Date: Mon, 30 May 2011 18:57:50 -0400
Subject: =?UTF-8?Q?Re=3A_Simple_URI_redirection_/_alias_in_Nginx=E2=80=A6?=
In-Reply-To: <70b8525468895548826d5775ed0bf6f6.NginxMailingListEnglish@forum.nginx.org>
References: <841ddad79482575f6eecec727b5e01de.NginxMailingListEnglish@forum.nginx.org>
	<70b8525468895548826d5775ed0bf6f6.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <a81d8f9dc9c4e7584e8bdffb211c0007.NginxMailingListEnglish@forum.nginx.org>

I meant this: 

   location  / {

Not "location = / {" of course. 

Thanks! 



-------------------------------------------------------
> I have a similar issue. 
> 
> Do we need separate location directives for each
> and every possible path, or can we do some
> "rewrite" inside location for the main / root
> folder? For example: 
> 
> location = / {
> 
>   # rewrite rules come here..
>   rewrite "^/adm$" "/wordpress/wp-login.php";
>   ....
>   fastcgi_pass 127.0.0.1:9000;
>   fastcgi_param SCRIPT_FILENAME
> /var/www/example.com/wordpress/wp-login.php
>   include /etc/nginx/fastcgi_params;
> 
> }
> 
> 
> 
> To me this is cleaner, but want to make sure that
> this is not the less desirable way? 
> 
> Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,202235,202431#msg-202431



From nginx-forum at nginx.us  Tue May 31 03:03:17 2011
From: nginx-forum at nginx.us (atipico)
Date: Mon, 30 May 2011 19:03:17 -0400
Subject: =?UTF-8?Q?Re=3A_Simple_URI_redirection_/_alias_in_Nginx=E2=80=A6?=
In-Reply-To: <20110530184709.GD87429@sysoev.ru>
References: <20110530184709.GD87429@sysoev.ru>
Message-ID: <6dbb1ac66c6cdbf93869ec96357ad27a.NginxMailingListEnglish@forum.nginx.org>

Thank you Igor. 

So what is the difference among doing what you told and this?
rewrite ^/adm$ /wp/wp-login.php;

Isn't it the case of usind the "alias"?

Thank you onc more,
Rogerio.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,202235,202432#msg-202432



From nginx-forum at nginx.us  Tue May 31 04:07:03 2011
From: nginx-forum at nginx.us (pk899)
Date: Mon, 30 May 2011 20:07:03 -0400
Subject: Formatting of nginx.conf -- any impact on performance?
Message-ID: <784b5294f522fb504e063c13c08fc9d7.NginxMailingListEnglish@forum.nginx.org>

Hi. Probably a silly question, but I have noticed that if I have a
properly tabbed/spaced config file, the performance of nginx goes down.
Is this a valid observation? 

Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,202451,202451#msg-202451



From gvimrc at gmail.com  Tue May 31 05:14:04 2011
From: gvimrc at gmail.com (gvim)
Date: Tue, 31 May 2011 02:14:04 +0100
Subject: Multiple server_name entries not working
Message-ID: <4DE440DC.9040605@gmail.com>

Nginx 1.0.3

I have a virtual host file with a simple config:

server {
     listen 8080;
     server_name   domain.com  www.domain.com;
     root   /var/www/html/dm2011/pub;
     access_log   /var/log/nginx/access.log main;

     location / {
       index index.htm index.html;
     }
}

Requests for 'domain.com' return the correct index page but requests for 'www.domain.com' return only the "Welcome to nginx!" page. I tried *.domain.com also as the 2nd name but the result was the same. There are no DNS issues with both versions of the domain.

gvim


From igor at sysoev.ru  Tue May 31 06:07:02 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 31 May 2011 06:07:02 +0400
Subject: Multiple server_name entries not working
In-Reply-To: <4DE440DC.9040605@gmail.com>
References: <4DE440DC.9040605@gmail.com>
Message-ID: <20110531020702.GA5992@sysoev.ru>

On Tue, May 31, 2011 at 02:14:04AM +0100, gvim wrote:
> Nginx 1.0.3
> 
> I have a virtual host file with a simple config:
> 
> server {
>      listen 8080;
>      server_name   domain.com  www.domain.com;
>      root   /var/www/html/dm2011/pub;
>      access_log   /var/log/nginx/access.log main;
> 
>      location / {
>        index index.htm index.html;
>      }
> }
> 
> Requests for 'domain.com' return the correct index page but requests for 'www.domain.com' return only the "Welcome to nginx!" page. I tried *.domain.com also as the 2nd name but the result was the same. There are no DNS issues with both versions of the domain.

If you set
      server_name   domain.com  www.domain.com;
what does "nginx -t" show ?


-- 
Igor Sysoev


From igor at sysoev.ru  Tue May 31 06:07:17 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 31 May 2011 06:07:17 +0400
Subject: Formatting of nginx.conf -- any impact on performance?
In-Reply-To: <784b5294f522fb504e063c13c08fc9d7.NginxMailingListEnglish@forum.nginx.org>
References: <784b5294f522fb504e063c13c08fc9d7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110531020717.GB5992@sysoev.ru>

On Mon, May 30, 2011 at 08:07:03PM -0400, pk899 wrote:
> Hi. Probably a silly question, but I have noticed that if I have a
> properly tabbed/spaced config file, the performance of nginx goes down.
> Is this a valid observation? 

No.


-- 
Igor Sysoev


From igor at sysoev.ru  Tue May 31 06:11:06 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 31 May 2011 06:11:06 +0400
Subject: =?UTF-8?Q?Re=3A_Simple_URI_redirection_/_alias_in_Nginx=E2=80=A6?=
In-Reply-To: <6dbb1ac66c6cdbf93869ec96357ad27a.NginxMailingListEnglish@forum.nginx.org>
References: <20110530184709.GD87429@sysoev.ru>
	<6dbb1ac66c6cdbf93869ec96357ad27a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110531021105.GC5992@sysoev.ru>

On Mon, May 30, 2011 at 07:03:17PM -0400, atipico wrote:
> Thank you Igor. 
> 
> So what is the difference among doing what you told and this?
> rewrite ^/adm$ /wp/wp-login.php;

Configuration maintenance. If you will not use rewrites and regex locations,
then you can easy change large configuration.

> Isn't it the case of usind the "alias"?

"alias" is for static files, but not scripts.


-- 
Igor Sysoev


From gvimrc at gmail.com  Tue May 31 06:34:19 2011
From: gvimrc at gmail.com (gvim)
Date: Tue, 31 May 2011 03:34:19 +0100
Subject: Multiple server_name entries not working
In-Reply-To: <20110531020702.GA5992@sysoev.ru>
References: <4DE440DC.9040605@gmail.com> <20110531020702.GA5992@sysoev.ru>
Message-ID: <4DE453AB.3090303@gmail.com>

On 31/05/2011 03:07, Igor Sysoev wrote:

> If you set server_name   domain.com  www.domain.com; what does "nginx
> -t" show ?
>

It gave a syntax OK response. It's mysteriously fixed now after retyping the line. Maybe my eyesight is going.

gvim
  



From david at somers-harris.com  Tue May 31 11:21:57 2011
From: david at somers-harris.com (David Somers Harris)
Date: Tue, 31 May 2011 16:21:57 +0900
Subject: HttpImageFilterModule: Crop with nonzero origin
In-Reply-To: <20110530095003.GL42265@mdounin.ru>
References: <BANLkTik5zeVtGp4kH=WrEfFVrA5pAjsWgQ@mail.gmail.com>
	<20110530095003.GL42265@mdounin.ru>
Message-ID: <BANLkTi=bBhomQCrQt5xjXtx_9RrDKh+GTw@mail.gmail.com>

Ah I see. Too bad.

When the Image Filter Module crops, does it crop out the right and bottom
ends only, or does it evenly crop the top, left, right and bottom out?

David
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110531/debaed05/attachment.html>

From igor at sysoev.ru  Tue May 31 11:40:19 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 31 May 2011 11:40:19 +0400
Subject: HttpImageFilterModule: Crop with nonzero origin
In-Reply-To: <BANLkTi=bBhomQCrQt5xjXtx_9RrDKh+GTw@mail.gmail.com>
References: <BANLkTik5zeVtGp4kH=WrEfFVrA5pAjsWgQ@mail.gmail.com>
	<20110530095003.GL42265@mdounin.ru>
	<BANLkTi=bBhomQCrQt5xjXtx_9RrDKh+GTw@mail.gmail.com>
Message-ID: <20110531074019.GA21552@sysoev.ru>

On Tue, May 31, 2011 at 04:21:57PM +0900, David Somers Harris wrote:
> Ah I see. Too bad.
> 
> When the Image Filter Module crops, does it crop out the right and bottom
> ends only, or does it evenly crop the top, left, right and bottom out?

It tries to keep as much as possible of an original image:
The image is scaled on one dimension and then is croped on other dimension.


-- 
Igor Sysoev


From nbubingo at gmail.com  Tue May 31 11:58:53 2011
From: nbubingo at gmail.com (=?GB2312?B?0qbOsLHz?=)
Date: Tue, 31 May 2011 15:58:53 +0800
Subject: NGINX loadbalancing ORACLE forms backends?
In-Reply-To: <4DE37F0F.97C0.005E.0@herold.at>
References: <4DDF9C640200005E0003480F@mx.herold.at>
	<BANLkTi=m1zbSLoP=3S-Q0gPxUbBJ39_P+w@mail.gmail.com>
	<4DE35D820200005E000348CF@mx.herold.at>
	<4DE37F0F.97C0.005E.0@herold.at>
Message-ID: <BANLkTi=M3OtdUpmqGqXg3eAzsSedgxc87w@mail.gmail.com>

Hi Dalibor,

This may be a configure sequence problem.  Do you put the upstream block
behind the server block? The upstream block should be put before the server
block. If you try other options, like "weight=10", there will be problem
too.

2011/5/30 Dalibor Jotanovic <Dalibor.Jotanovic at herold.at>

>  Hi again,
>
> iirc that jvm route plugin is not supported by nginx. Anyway if somebody is
> having the same issue i had with that plugin and the mentioned srun config
> option, here is the *possible* solution to it:
>
> Set the *jvm_route $cookie JESSIONID;*  *before *the server and srun
> definitons. Otherwise the plugin is not loaded and the srun option is
> unknown i guess...
>
> Cheers,
> DJ
>
>
> >>> "Dalibor Jotanovic" <Dalibor.Jotanovic at herold.at> 5/30/2011 9:04 >>>
> Hi,
>
> i am actually using this module at the moment, but the documentated
> config option "srun...."  doesn't seem to work in the latest nginx version.
> Does the option work on your side? Don't you get any errors?
>
> Cheers,
> DJ
>
> >>> ???<nbubingo at gmail.com> 5/27/2011 1:20 >>>
> See also this module: http://code.google.com/p/nginx-upstream-jvm-route/
>
>  <http://code.google.com/p/nginx-upstream-jvm-route/>This module achieves
> session stickiness with the session cookie. If the session is not in the
> cookie or URL, the module will be a normal Round-Robin upstream module.
>
> 2011/5/27 Dalibor Jotanovic <Dalibor.Jotanovic at herold.at>
>
>>  Hi,
>>  does anybody have a working setup with NGINX loadbalancing ORACLE FORMS
>> backends?
>> I'm interested in setting up an environment with JSESSION cookie based
>> sticky sessions.
>>  Regards,
>> DJ
>>
>>  <http://www.herold.at>
>> <http://www.herold.at>
>>
>> HEROLD Business Data GmbH
>> Guntramsdorfer Stra?e 105
>> A-2340 M?dling
>>
>> FN 233171z
>> Landesgericht Wiener Neustadt
>>
>> Besuchen Sie uns online und mobil auf *www.herold.at*<http://www.herold.at>
>> !
>>
>> Weitere Informationen zu unseren Produkten finden Sie unter:
>> *http://mobile.herold.at* <http://mobile.herold.at>
>> *http://daten.herold.at * <http://daten.herold.at>
>>
>> <http://www.facebook.com/derherold>Werden Sie Fan von HEROLD auf
>> Facebook!
>>
>>
>> Bitte beachten Sie auch unsere Tochter- und Partner-Unternehmen:
>> *www.clearsense.at * <http://www.clearsense.at>
>> *www.tupalo.com * <http://www.tupalo.com>
>> *www.aboutmedia.at* <http://www.aboutmedia.at>* <http://www.tupalo.com>*
>> *www.love.at* <http://www.love.at>* <http://www.tupalo.com>*
>> *www.reise-hero.at* <http://www.reise-hero.at>* <http://www.tupalo.com>*
>> *www.urlauburlaub.at* <http://www.urlauburlaub.at>*
>> <http://www.tupalo.com>*
>> <http://www.tupalo.com>
>> Diese E-Mail kann vertrauliche und/oder rechtlich gesch?tzte Informationen
>> enthalten. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail
>> irrt?mlich erhalten haben, informieren Sie bitte sofort den Absender und
>> vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie die unbefugte
>> Weitergabe dieser Mail sind nicht gestattet.
>>
>> This e-mail may contain confidential and/or privileged information. If you
>> are not the intended recipient (or have received this e-mail in error)
>> please notify the sender immediately and delete this e-mail. Any
>> unauthorized copying, disclosure or distribution of the material in this
>> e-mail is strictly forbidden.
>>
>> _______________________________________________
>> nginx mailing list
>> nginx at nginx.org
>> http://nginx.org/mailman/listinfo/nginx
>>
>>
>
>
> Click here<https://www.mailcontrol.com/sr/t91yjfYfe5jTndxI!oX7Uh5Cv+dgaYj8ylv+aQFyzJM2yPGdnNdCJu57SCsCe4JD1GQ2ktvKI6YpdY6WjzZSeg==>to report this email as spam.
>
>  <http://www.herold.at>
> <http://www.herold.at>
>
> HEROLD Business Data GmbH
> Guntramsdorfer Stra?e 105
> A-2340 M?dling
>
> FN 233171z
> Landesgericht Wiener Neustadt
>
> Besuchen Sie uns online und mobil auf *www.herold.at*<http://www.herold.at>
> !
>
> Weitere Informationen zu unseren Produkten finden Sie unter:
> *http://mobile.herold.at* <http://mobile.herold.at>
> *http://daten.herold.at * <http://daten.herold.at>
>
> <http://www.facebook.com/derherold>Werden Sie Fan von HEROLD auf Facebook!
>
>
> Bitte beachten Sie auch unsere Tochter- und Partner-Unternehmen:
> *www.clearsense.at * <http://www.clearsense.at>
> *www.tupalo.com * <http://www.tupalo.com>
> *www.aboutmedia.at* <http://www.aboutmedia.at>* <http://www.tupalo.com>*
> *www.love.at* <http://www.love.at>* <http://www.tupalo.com>*
> *www.reise-hero.at* <http://www.reise-hero.at>* <http://www.tupalo.com>*
> *www.urlauburlaub.at* <http://www.urlauburlaub.at>*
> <http://www.tupalo.com>*
> <http://www.tupalo.com>
> Diese E-Mail kann vertrauliche und/oder rechtlich gesch?tzte Informationen
> enthalten. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail
> irrt?mlich erhalten haben, informieren Sie bitte sofort den Absender und
> vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie die unbefugte
> Weitergabe dieser Mail sind nicht gestattet.
>
> This e-mail may contain confidential and/or privileged information. If you
> are not the intended recipient (or have received this e-mail in error)
> please notify the sender immediately and delete this e-mail. Any
> unauthorized copying, disclosure or distribution of the material in this
> e-mail is strictly forbidden.
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110531/4baacb4b/attachment-0001.html>

From nginx-forum at nginx.us  Tue May 31 15:20:34 2011
From: nginx-forum at nginx.us (csg)
Date: Tue, 31 May 2011 07:20:34 -0400
Subject: Accessing binary blob data in MySQL backend
Message-ID: <547a0936722abfbca78f1b68df1588dd.NginxMailingListEnglish@forum.nginx.org>

Hi!

I am curious if is feasible to have Nginx deliver binary blob data which
is currently stored in a MySQL database. I was loosely following the
examples mentioned in agentzh's blog [1] which work all fine for me with
non-binary data. But as JSON is used for intermediate serialization
which is not binary safe I was wondering if there are other ways (or
even modules) that access the raw payload data from its RDS
representation.

In my scenario I'm trying to access images data like JPEGs stored as
blobs in MySQL causing the JSON parser to fail of course

content_by_lua aborted: runtime error: InvalidJSONInput: lexical error:
invalid bytes in UTF8 string.

My current software stack consists of CentOS 5.6/x86_64, Nignx 1.0.3
(plus simpl-ngx_devel_kit 0.2.17, chaoslawful-lua-nginx-module 0.1.5,
chaoslawful-drizzle-nginx-module 0.0.14, agentzh-rds-json-nginx-module
0.10), Lua 5.1.4, LuaRocks 2.0.4.1, Lua YAJL 2.0, YAJL 2.0.1, Drizzle7
libraries 2011.04.15 and MySQL database 5.1.57.

[1]
http://agentzh.blogspot.com/2011/02/ngxlua-v015-ability-to-capture-multiple_6605.html

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,202577,202577#msg-202577



From nginx-forum at nginx.us  Tue May 31 15:50:52 2011
From: nginx-forum at nginx.us (Ensiferous)
Date: Tue, 31 May 2011 07:50:52 -0400
Subject: Multiple server_name entries not working
In-Reply-To: <4DE440DC.9040605@gmail.com>
References: <4DE440DC.9040605@gmail.com>
Message-ID: <f7694f5051625fcf65d442e4904586c5.NginxMailingListEnglish@forum.nginx.org>

Sounds like it was a browser caching issue. Browsers cache aggressively
so always test with curl when you're making changes.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,202461,202582#msg-202582



From mp3geek at gmail.com  Tue May 31 17:04:34 2011
From: mp3geek at gmail.com (Ryan B)
Date: Wed, 1 Jun 2011 01:04:34 +1200
Subject: Tweaking nginx workers nice values
Message-ID: <BANLkTinY6Tf47QNBHKZJ3_LyUtc17UvAvA@mail.gmail.com>

Is it possible to set the nice values for nginx worker processes, the
main processs can be tweaked easily.. but I'd like to tweak the worker
processes too..

Thanks.


From mp3geek at gmail.com  Tue May 31 17:27:07 2011
From: mp3geek at gmail.com (Ryan B)
Date: Wed, 1 Jun 2011 01:27:07 +1200
Subject: Tweaking nginx workers nice values
In-Reply-To: <BANLkTinY6Tf47QNBHKZJ3_LyUtc17UvAvA@mail.gmail.com>
References: <BANLkTinY6Tf47QNBHKZJ3_LyUtc17UvAvA@mail.gmail.com>
Message-ID: <BANLkTi=RPcqYDJcZuF86N4e7HyZWxrHCLQ@mail.gmail.com>

ah, discovered worker_priority  variable :)

On Wed, Jun 1, 2011 at 1:04 AM, Ryan B <mp3geek at gmail.com> wrote:
> Is it possible to set the nice values for nginx worker processes, the
> main processs can be tweaked easily.. but I'd like to tweak the worker
> processes too..
>
> Thanks.
>


From igor at sysoev.ru  Tue May 31 17:57:35 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 31 May 2011 17:57:35 +0400
Subject: Accessing binary blob data in MySQL backend
In-Reply-To: <547a0936722abfbca78f1b68df1588dd.NginxMailingListEnglish@forum.nginx.org>
References: <547a0936722abfbca78f1b68df1588dd.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110531135734.GA56723@sysoev.ru>

On Tue, May 31, 2011 at 07:20:34AM -0400, csg wrote:
> Hi!
> 
> I am curious if is feasible to have Nginx deliver binary blob data which
> is currently stored in a MySQL database. I was loosely following the
> examples mentioned in agentzh's blog [1] which work all fine for me with
> non-binary data. But as JSON is used for intermediate serialization
> which is not binary safe I was wondering if there are other ways (or
> even modules) that access the raw payload data from its RDS
> representation.
> 
> In my scenario I'm trying to access images data like JPEGs stored as
> blobs in MySQL causing the JSON parser to fail of course

I can not say anuthing about subject, I'm just interesting are any
advanatges to store images in MySQL instead of file system ?


-- 
Igor Sysoev


From mat999 at gmail.com  Tue May 31 18:06:12 2011
From: mat999 at gmail.com (SplitIce)
Date: Wed, 1 Jun 2011 00:06:12 +1000
Subject: Accessing binary blob data in MySQL backend
In-Reply-To: <20110531135734.GA56723@sysoev.ru>
References: <547a0936722abfbca78f1b68df1588dd.NginxMailingListEnglish@forum.nginx.org>
	<20110531135734.GA56723@sysoev.ru>
Message-ID: <BANLkTimuskfT4VcEe3BAX2mX-qHEnRv0LA@mail.gmail.com>

ease of management is the main positive, I recently moved away from it
myself to a filesystem based solution for one of my clients. The largest
disadvantage is the size of the table. Oh and JSON can be binary safe if
implmented correctly with correct escape sequences, yes?

On Tue, May 31, 2011 at 11:57 PM, Igor Sysoev <igor at sysoev.ru> wrote:

> On Tue, May 31, 2011 at 07:20:34AM -0400, csg wrote:
> > Hi!
> >
> > I am curious if is feasible to have Nginx deliver binary blob data which
> > is currently stored in a MySQL database. I was loosely following the
> > examples mentioned in agentzh's blog [1] which work all fine for me with
> > non-binary data. But as JSON is used for intermediate serialization
> > which is not binary safe I was wondering if there are other ways (or
> > even modules) that access the raw payload data from its RDS
> > representation.
> >
> > In my scenario I'm trying to access images data like JPEGs stored as
> > blobs in MySQL causing the JSON parser to fail of course
>
> I can not say anuthing about subject, I'm just interesting are any
> advanatges to store images in MySQL instead of file system ?
>
>
> --
> Igor Sysoev
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>



-- 
Warez Scene <http://thewarezscene.org> Free Rapidshare
Downloads<http://www.nexusddl.com>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110601/5b991316/attachment.html>

From igor at sysoev.ru  Tue May 31 18:25:44 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 31 May 2011 18:25:44 +0400
Subject: Tweaking nginx workers nice values
In-Reply-To: <BANLkTinY6Tf47QNBHKZJ3_LyUtc17UvAvA@mail.gmail.com>
References: <BANLkTinY6Tf47QNBHKZJ3_LyUtc17UvAvA@mail.gmail.com>
Message-ID: <20110531142544.GB56723@sysoev.ru>

On Wed, Jun 01, 2011 at 01:04:34AM +1200, Ryan B wrote:
> Is it possible to set the nice values for nginx worker processes, the
> main processs can be tweaked easily.. but I'd like to tweak the worker
> processes too..

worker_priority  -10;


-- 
Igor Sysoev


From igor at sysoev.ru  Tue May 31 18:27:29 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 31 May 2011 18:27:29 +0400
Subject: Tweaking nginx workers nice values
In-Reply-To: <BANLkTi=RPcqYDJcZuF86N4e7HyZWxrHCLQ@mail.gmail.com>
References: <BANLkTinY6Tf47QNBHKZJ3_LyUtc17UvAvA@mail.gmail.com>
	<BANLkTi=RPcqYDJcZuF86N4e7HyZWxrHCLQ@mail.gmail.com>
Message-ID: <20110531142729.GC56723@sysoev.ru>

On Wed, Jun 01, 2011 at 01:27:07AM +1200, Ryan B wrote:
> ah, discovered worker_priority  variable :)

This is not variable. This is a directive.


-- 
Igor Sysoev


From igor at sysoev.ru  Tue May 31 18:33:18 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 31 May 2011 18:33:18 +0400
Subject: Accessing binary blob data in MySQL backend
In-Reply-To: <BANLkTimuskfT4VcEe3BAX2mX-qHEnRv0LA@mail.gmail.com>
References: <547a0936722abfbca78f1b68df1588dd.NginxMailingListEnglish@forum.nginx.org>
	<20110531135734.GA56723@sysoev.ru>
	<BANLkTimuskfT4VcEe3BAX2mX-qHEnRv0LA@mail.gmail.com>
Message-ID: <20110531143318.GD56723@sysoev.ru>

On Wed, Jun 01, 2011 at 12:06:12AM +1000, SplitIce wrote:
> ease of management is the main positive, I recently moved away from it

What do you mean by "ease of management". As for me there is no easier
management than file system one. You have a lot of tools -  find, grep, etc.

> myself to a filesystem based solution for one of my clients. The largest
> disadvantage is the size of the table. Oh and JSON can be binary safe if
> implmented correctly with correct escape sequences, yes?
> 
> On Tue, May 31, 2011 at 11:57 PM, Igor Sysoev <igor at sysoev.ru> wrote:
> 
> > On Tue, May 31, 2011 at 07:20:34AM -0400, csg wrote:
> > > Hi!
> > >
> > > I am curious if is feasible to have Nginx deliver binary blob data which
> > > is currently stored in a MySQL database. I was loosely following the
> > > examples mentioned in agentzh's blog [1] which work all fine for me with
> > > non-binary data. But as JSON is used for intermediate serialization
> > > which is not binary safe I was wondering if there are other ways (or
> > > even modules) that access the raw payload data from its RDS
> > > representation.
> > >
> > > In my scenario I'm trying to access images data like JPEGs stored as
> > > blobs in MySQL causing the JSON parser to fail of course
> >
> > I can not say anuthing about subject, I'm just interesting are any
> > advanatges to store images in MySQL instead of file system ?


-- 
Igor Sysoev


From mat999 at gmail.com  Tue May 31 18:38:33 2011
From: mat999 at gmail.com (SplitIce)
Date: Wed, 1 Jun 2011 00:38:33 +1000
Subject: Accessing binary blob data in MySQL backend
In-Reply-To: <20110531143318.GD56723@sysoev.ru>
References: <547a0936722abfbca78f1b68df1588dd.NginxMailingListEnglish@forum.nginx.org>
	<20110531135734.GA56723@sysoev.ru>
	<BANLkTimuskfT4VcEe3BAX2mX-qHEnRv0LA@mail.gmail.com>
	<20110531143318.GD56723@sysoev.ru>
Message-ID: <BANLkTinYPD2aNuLPgN=n8zjfy7-56JQVjA@mail.gmail.com>

DELETE FROM images WHERE (uploaded_user='user1' OR uploaded_user='user2')
AND uploaded_date > 123 AND gallery_id=0

SQL makes querying and management easy, thats why some people choose to
store their data in mysql BLOBs instead of on the filesystem.



On Wed, Jun 1, 2011 at 12:33 AM, Igor Sysoev <igor at sysoev.ru> wrote:

> On Wed, Jun 01, 2011 at 12:06:12AM +1000, SplitIce wrote:
> > ease of management is the main positive, I recently moved away from it
>
> What do you mean by "ease of management". As for me there is no easier
> management than file system one. You have a lot of tools -  find, grep,
> etc.
>
> > myself to a filesystem based solution for one of my clients. The largest
> > disadvantage is the size of the table. Oh and JSON can be binary safe if
> > implmented correctly with correct escape sequences, yes?
> >
> > On Tue, May 31, 2011 at 11:57 PM, Igor Sysoev <igor at sysoev.ru> wrote:
> >
> > > On Tue, May 31, 2011 at 07:20:34AM -0400, csg wrote:
> > > > Hi!
> > > >
> > > > I am curious if is feasible to have Nginx deliver binary blob data
> which
> > > > is currently stored in a MySQL database. I was loosely following the
> > > > examples mentioned in agentzh's blog [1] which work all fine for me
> with
> > > > non-binary data. But as JSON is used for intermediate serialization
> > > > which is not binary safe I was wondering if there are other ways (or
> > > > even modules) that access the raw payload data from its RDS
> > > > representation.
> > > >
> > > > In my scenario I'm trying to access images data like JPEGs stored as
> > > > blobs in MySQL causing the JSON parser to fail of course
> > >
> > > I can not say anuthing about subject, I'm just interesting are any
> > > advanatges to store images in MySQL instead of file system ?
>
>
> --
> Igor Sysoev
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>



-- 
Warez Scene <http://thewarezscene.org> Free Rapidshare
Downloads<http://www.nexusddl.com>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nginx.org/pipermail/nginx/attachments/20110601/7accbaac/attachment-0001.html>

From nginx-forum at nginx.us  Tue May 31 18:42:31 2011
From: nginx-forum at nginx.us (csg)
Date: Tue, 31 May 2011 10:42:31 -0400
Subject: Accessing binary blob data in MySQL backend
In-Reply-To: <20110531135734.GA56723@sysoev.ru>
References: <20110531135734.GA56723@sysoev.ru>
Message-ID: <cc7521b20d30183984c16d639cb76df7.NginxMailingListEnglish@forum.nginx.org>

Igor Sysoev Wrote:
-------------------------------------------------------
 
> I can not say anuthing about subject, I'm just
> interesting are any
> advanatges to store images in MySQL instead of
> file system ?

First of all quite a lot of backend processes and application code
relies on it. Those interfaces cannot be changed easily. But on the
other hand for the ease of handling the data. For redundancy it is much
easier and more performant to do MySQL replication instead of file
system based approaches. With millions of files even when using deep
hashed subdirectory trees a full rsync(1) run would take several hours
up to half a day to perform a sync. Even when using tools utilizing the
inotify(7) framework like lsyncd(1) [1] you have to run a full rsync(1)
run from time (e.g., reboots due to kernel updates, inotify event
overflow, etc.). With data backup you would face the same issues, too.
As company policies discourage the use of AGPL-licenses software,
MongoDB/gridfs is also not an option. Our current solution is Apache 2.2
MPM worker/Perl-based but on the long term this does not scale too well
anymore, that's why we are looking for alternatives.

[1] http://code.google.com/p/lsyncd/

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,202577,202648#msg-202648



From nginx-forum at nginx.us  Tue May 31 18:48:50 2011
From: nginx-forum at nginx.us (csg)
Date: Tue, 31 May 2011 10:48:50 -0400
Subject: Accessing binary blob data in MySQL backend
In-Reply-To: <BANLkTimuskfT4VcEe3BAX2mX-qHEnRv0LA@mail.gmail.com>
References: <BANLkTimuskfT4VcEe3BAX2mX-qHEnRv0LA@mail.gmail.com>
Message-ID: <016b6e4deb2ac124dbecc7857523eaed.NginxMailingListEnglish@forum.nginx.org>

SplitIce Wrote:
-------------------------------------------------------
> The largest
> disadvantage is the size of the table.

This can be easily avoided by sharding the data into many tables,
accessing it per modulo operation on the numerical key.

> Oh and JSON
> can be binary safe if
> implmented correctly with correct escape
> sequences, yes?

But this would require -- if I am not mistaken -- to touch all data end
encode it with Base64 (or similar) which gives you computation and data
capacity overhead. Additionally this would require changes to
application which we are trying to avoid. Some component inbetwen like
MySQL Proxy might be able to workaround this, but this increases latency
which is also not desired.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,202577,202651#msg-202651



From anil at saog.net  Tue May 31 18:52:02 2011
From: anil at saog.net (=?UTF-8?B?QW7EsWwgw4dldGlu?=)
Date: Tue, 31 May 2011 17:52:02 +0300
Subject: Accessing binary blob data in MySQL backend
In-Reply-To: <cc7521b20d30183984c16d639cb76df7.NginxMailingListEnglish@forum.nginx.org>
References: <20110531135734.GA56723@sysoev.ru>
	<cc7521b20d30183984c16d639cb76df7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4DE50092.9060400@saog.net>

You may try "fuse" (http://fuse.sourceforge.net/), implement a tiny file 
system easily which uses mysql as data source and acts as a real file 
system.

On 5/31/2011 5:42 PM, csg wrote:
> Igor Sysoev Wrote:
> -------------------------------------------------------
>
>> I can not say anuthing about subject, I'm just
>> interesting are any
>> advanatges to store images in MySQL instead of
>> file system ?
> First of all quite a lot of backend processes and application code
> relies on it. Those interfaces cannot be changed easily. But on the
> other hand for the ease of handling the data. For redundancy it is much
> easier and more performant to do MySQL replication instead of file
> system based approaches. With millions of files even when using deep
> hashed subdirectory trees a full rsync(1) run would take several hours
> up to half a day to perform a sync. Even when using tools utilizing the
> inotify(7) framework like lsyncd(1) [1] you have to run a full rsync(1)
> run from time (e.g., reboots due to kernel updates, inotify event
> overflow, etc.). With data backup you would face the same issues, too.
> As company policies discourage the use of AGPL-licenses software,
> MongoDB/gridfs is also not an option. Our current solution is Apache 2.2
> MPM worker/Perl-based but on the long term this does not scale too well
> anymore, that's why we are looking for alternatives.
>
> [1] http://code.google.com/p/lsyncd/
>
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,202577,202648#msg-202648
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://nginx.org/mailman/listinfo/nginx
>



From nginx-forum at nginx.us  Tue May 31 19:05:09 2011
From: nginx-forum at nginx.us (csg)
Date: Tue, 31 May 2011 11:05:09 -0400
Subject: Accessing binary blob data in MySQL backend
In-Reply-To: <4DE50092.9060400@saog.net>
References: <4DE50092.9060400@saog.net>
Message-ID: <004b4559b17a8f3403da518681738b0b.NginxMailingListEnglish@forum.nginx.org>

mow Wrote:
-------------------------------------------------------
> You may try "fuse" (http://fuse.sourceforge.net/),
> implement a tiny file 
> system easily which uses mysql as data source and
> acts as a real file 
> system.

Correct me if I am wrong, but I reckon that context switches between
kernel and a userland filesystem would be detrimental to the
performance.Performance is the key ;-)

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,202577,202655#msg-202655



From igor at sysoev.ru  Tue May 31 19:08:03 2011
From: igor at sysoev.ru (Igor Sysoev)
Date: Tue, 31 May 2011 19:08:03 +0400
Subject: Accessing binary blob data in MySQL backend
In-Reply-To: <004b4559b17a8f3403da518681738b0b.NginxMailingListEnglish@forum.nginx.org>
References: <4DE50092.9060400@saog.net>
	<004b4559b17a8f3403da518681738b0b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20110531150803.GF56723@sysoev.ru>

On Tue, May 31, 2011 at 11:05:09AM -0400, csg wrote:
> mow Wrote:
> -------------------------------------------------------
> > You may try "fuse" (http://fuse.sourceforge.net/),
> > implement a tiny file 
> > system easily which uses mysql as data source and
> > acts as a real file 
> > system.
> 
> Correct me if I am wrong, but I reckon that context switches between
> kernel and a userland filesystem would be detrimental to the
> performance.Performance is the key ;-)

If perfomance is the key, you should use a file system.
Or you believe that reading from MySQL is magic zero transportation ? :)


-- 
Igor Sysoev


From nginx-forum at nginx.us  Tue May 31 19:24:41 2011
From: nginx-forum at nginx.us (csg)
Date: Tue, 31 May 2011 11:24:41 -0400
Subject: Accessing binary blob data in MySQL backend
In-Reply-To: <20110531150803.GF56723@sysoev.ru>
References: <20110531150803.GF56723@sysoev.ru>
Message-ID: <4cdcc593ca373de3f2c5398ee3f4b03d.NginxMailingListEnglish@forum.nginx.org>

Igor Sysoev Wrote:
-------------------------------------------------------
> If perfomance is the key, you should use a file
> system.
> Or you believe that reading from MySQL is magic
> zero transportation ? :)

For sure not, and I didn't say so. But instead of bringing yet another
software component into play and entering a semi-flameware about the
pros and cons to store blobs in databases (there are already so many on
the net, and I am looking for information I couldn't find on the net) I
had rather hoped to get some backgr