From mdounin at mdounin.ru  Sun Jan  1 00:34:57 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sun, 1 Jan 2012 04:34:57 +0400
Subject: Is nginx vulnerable to the Hash Table Vulnerability (n.runs AG)?
In-Reply-To: <CAEHYtVjfNoA6uYQE+DmOFEzj05znaMoNmUGRa6FtOTsD2EOJQQ@mail.gmail.com>
References: <CAEHYtVjfNoA6uYQE+DmOFEzj05znaMoNmUGRa6FtOTsD2EOJQQ@mail.gmail.com>
Message-ID: <20120101003457.GY67687@mdounin.ru>

Hello!

On Sat, Dec 31, 2011 at 11:37:39AM -0700, Justin Hart wrote:

> http://www.securityweek.com/hash-table-collision-attacks-could-trigger-ddos-massive-scale
> 
> Without going through the way nginx parses an incoming request, I'm unsure
> if nginx isn't vulnerable to this, because of the availability to grab the
> value of a GET parameter via
> http://wiki.nginx.org/HttpCoreModule#.24arg_PARAMETER.  My hope is that
> especially if an $arg_PARAMETER isn't used in the config, it is not
> vulnerable because it wouldn't even attempt to parse the parameters, but I
> can't be sure.
> 
> Can anyone speak to this?

It's not vulnerable even if $arg_* is used.

Maxim Dounin


From agentzh at gmail.com  Sun Jan  1 05:54:08 2012
From: agentzh at gmail.com (agentzh)
Date: Sun, 1 Jan 2012 13:54:08 +0800
Subject: Is nginx vulnerable to the Hash Table Vulnerability (n.runs AG)?
In-Reply-To: <CAEHYtVjfNoA6uYQE+DmOFEzj05znaMoNmUGRa6FtOTsD2EOJQQ@mail.gmail.com>
References: <CAEHYtVjfNoA6uYQE+DmOFEzj05znaMoNmUGRa6FtOTsD2EOJQQ@mail.gmail.com>
Message-ID: <CAB4Tn6OU5h3m5GRkZUTynoixJ9MedyXvduhXh4kCNQvkzrc3TQ@mail.gmail.com>

On Sun, Jan 1, 2012 at 2:37 AM, Justin Hart <onyxraven at gmail.com> wrote:
> http://www.securityweek.com/hash-table-collision-attacks-could-trigger-ddos-massive-scale
>
> Without going through the way nginx parses an incoming request, I'm unsure
> if nginx isn't vulnerable to this, because of the availability to grab the
> value of a GET parameter
> via?http://wiki.nginx.org/HttpCoreModule#.24arg_PARAMETER. ?My hope is that
> especially if an $arg_PARAMETER isn't used in the config, it is not
> vulnerable because it wouldn't even attempt to parse the parameters, but I
> can't be sure.
>

Well, the $arg_PARAMETER variable is not implemented with hash tables
at all ;) It scans the URI query string at every invocation :)

Regards,
-agentzh


From onyxraven at gmail.com  Sun Jan  1 05:58:48 2012
From: onyxraven at gmail.com (Justin Hart)
Date: Sat, 31 Dec 2011 22:58:48 -0700
Subject: Is nginx vulnerable to the Hash Table Vulnerability (n.runs AG)?
In-Reply-To: <CAB4Tn6OU5h3m5GRkZUTynoixJ9MedyXvduhXh4kCNQvkzrc3TQ@mail.gmail.com>
References: <CAEHYtVjfNoA6uYQE+DmOFEzj05znaMoNmUGRa6FtOTsD2EOJQQ@mail.gmail.com>
	<CAB4Tn6OU5h3m5GRkZUTynoixJ9MedyXvduhXh4kCNQvkzrc3TQ@mail.gmail.com>
Message-ID: <1285211291852136747@unknownmsgid>

Thank you for the confirmation - I read through the parts of code in
question but wanted to get a second opinion.

How about the lua and/or the perl modules?  It looks as if they are
using the nginx functions?

Sent from my iPhone

On Dec 31, 2011, at 10:54 PM, agentzh <agentzh at gmail.com> wrote:

> On Sun, Jan 1, 2012 at 2:37 AM, Justin Hart <onyxraven at gmail.com> wrote:
>> http://www.securityweek.com/hash-table-collision-attacks-could-trigger-ddos-massive-scale
>>
>> Without going through the way nginx parses an incoming request, I'm unsure
>> if nginx isn't vulnerable to this, because of the availability to grab the
>> value of a GET parameter
>> via http://wiki.nginx.org/HttpCoreModule#.24arg_PARAMETER.  My hope is that
>> especially if an $arg_PARAMETER isn't used in the config, it is not
>> vulnerable because it wouldn't even attempt to parse the parameters, but I
>> can't be sure.
>>
>
> Well, the $arg_PARAMETER variable is not implemented with hash tables
> at all ;) It scans the URI query string at every invocation :)
>
> Regards,
> -agentzh
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From agentzh at gmail.com  Sun Jan  1 14:20:59 2012
From: agentzh at gmail.com (agentzh)
Date: Sun, 1 Jan 2012 22:20:59 +0800
Subject: Is nginx vulnerable to the Hash Table Vulnerability (n.runs AG)?
In-Reply-To: <1285211291852136747@unknownmsgid>
References: <CAEHYtVjfNoA6uYQE+DmOFEzj05znaMoNmUGRa6FtOTsD2EOJQQ@mail.gmail.com>
	<CAB4Tn6OU5h3m5GRkZUTynoixJ9MedyXvduhXh4kCNQvkzrc3TQ@mail.gmail.com>
	<1285211291852136747@unknownmsgid>
Message-ID: <CAB4Tn6PT9ZdEwt3eKzfXYS=wyT-A120RH7R_QgDrs9=a9DVi_A@mail.gmail.com>

On Sun, Jan 1, 2012 at 1:58 PM, Justin Hart <onyxraven at gmail.com> wrote:
> Thank you for the confirmation - I read through the parts of code in
> question but wanted to get a second opinion.
>
> How about the lua and/or the perl modules? ?It looks as if they are
> using the nginx functions?
>

The current released versions of ngx_lua does have this vulnerability
in its ngx.req.get_uri_args() and ngx.req.get_post_args() functions.
I've already worked out a patch for these two functions in ngx_lua's
git max-args branch here:

    https://github.com/chaoslawful/lua-nginx-module/commit/75876

With this patch, both of these functions will only parse 100 query
args at most. And one can specify a custom maximum number of args
parsed with an optional function argument (default to 100) and
enforcing unlimited parsing by specifying a zero number.

This patch (as well as this branch) will be merged into the master
branch in 3 Jan.

Best,
-agentzh


From nginx at nginxuser.net  Sun Jan  1 14:31:46 2012
From: nginx at nginxuser.net (Nginx User)
Date: Sun, 1 Jan 2012 17:31:46 +0300
Subject: Is nginx vulnerable to the Hash Table Vulnerability (n.runs AG)?
In-Reply-To: <CAB4Tn6PT9ZdEwt3eKzfXYS=wyT-A120RH7R_QgDrs9=a9DVi_A@mail.gmail.com>
References: <CAEHYtVjfNoA6uYQE+DmOFEzj05znaMoNmUGRa6FtOTsD2EOJQQ@mail.gmail.com>
	<CAB4Tn6OU5h3m5GRkZUTynoixJ9MedyXvduhXh4kCNQvkzrc3TQ@mail.gmail.com>
	<1285211291852136747@unknownmsgid>
	<CAB4Tn6PT9ZdEwt3eKzfXYS=wyT-A120RH7R_QgDrs9=a9DVi_A@mail.gmail.com>
Message-ID: <CAN=WzV3-Y=k6qPrspks-PwkLj1GJAtSMoCenN=BhUWqHnUCUqw@mail.gmail.com>

On 1 January 2012 17:20, agentzh <agentzh at gmail.com> wrote:
> On Sun, Jan 1, 2012 at 1:58 PM, Justin Hart <onyxraven at gmail.com> wrote:
>> Thank you for the confirmation - I read through the parts of code in
>> question but wanted to get a second opinion.
>>
>> How about the lua and/or the perl modules? ?It looks as if they are
>> using the nginx functions?
>>
>
> The current released versions of ngx_lua does have this vulnerability
> in its ngx.req.get_uri_args() and ngx.req.get_post_args() functions.
> I've already worked out a patch for these two functions in ngx_lua's
> git max-args branch here:
>
> ? ?https://github.com/chaoslawful/lua-nginx-module/commit/75876
>
> With this patch, both of these functions will only parse 100 query
> args at most. And one can specify a custom maximum number of args
> parsed with an optional function argument (default to 100) and
> enforcing unlimited parsing by specifying a zero number.
>
> This patch (as well as this branch) will be merged into the master
> branch in 3 Jan.

It would probably be a good idea at that point, to finally make a
release of v0.3.1 of the ngx_lua module as with about 45 "Release
Candidates", it must already hold some record :)


From osa at FreeBSD.org.ru  Sun Jan  1 18:38:09 2012
From: osa at FreeBSD.org.ru (Sergey A. Osokin)
Date: Sun, 1 Jan 2012 22:38:09 +0400
Subject: Is nginx vulnerable to the Hash Table Vulnerability (n.runs AG)?
In-Reply-To: <CAN=WzV3-Y=k6qPrspks-PwkLj1GJAtSMoCenN=BhUWqHnUCUqw@mail.gmail.com>
References: <CAEHYtVjfNoA6uYQE+DmOFEzj05znaMoNmUGRa6FtOTsD2EOJQQ@mail.gmail.com>
	<CAB4Tn6OU5h3m5GRkZUTynoixJ9MedyXvduhXh4kCNQvkzrc3TQ@mail.gmail.com>
	<1285211291852136747@unknownmsgid>
	<CAB4Tn6PT9ZdEwt3eKzfXYS=wyT-A120RH7R_QgDrs9=a9DVi_A@mail.gmail.com>
	<CAN=WzV3-Y=k6qPrspks-PwkLj1GJAtSMoCenN=BhUWqHnUCUqw@mail.gmail.com>
Message-ID: <20120101183809.GB82696@FreeBSD.org.ru>

On Sun, Jan 01, 2012 at 05:31:46PM +0300, Nginx User wrote:
> On 1 January 2012 17:20, agentzh <agentzh at gmail.com> wrote:
> > On Sun, Jan 1, 2012 at 1:58 PM, Justin Hart <onyxraven at gmail.com> wrote:
> >> Thank you for the confirmation - I read through the parts of code in
> >> question but wanted to get a second opinion.
> >>
> >> How about the lua and/or the perl modules? ?It looks as if they are
> >> using the nginx functions?
> >
> > The current released versions of ngx_lua does have this vulnerability
> > in its ngx.req.get_uri_args() and ngx.req.get_post_args() functions.
> > I've already worked out a patch for these two functions in ngx_lua's
> > git max-args branch here:
> >
> > ? ?https://github.com/chaoslawful/lua-nginx-module/commit/75876
> >
> > With this patch, both of these functions will only parse 100 query
> > args at most. And one can specify a custom maximum number of args
> > parsed with an optional function argument (default to 100) and
> > enforcing unlimited parsing by specifying a zero number.
> >
> > This patch (as well as this branch) will be merged into the master
> > branch in 3 Jan.
> 
> It would probably be a good idea at that point, to finally make a
> release of v0.3.1 of the ngx_lua module as with about 45 "Release
> Candidates", it must already hold some record :)

+1.

-- 
Sergey A. Osokin
osa at FreeBSD.ORG.ru
osa at FreeBSD.ORG


From nginx-forum at nginx.us  Sun Jan  1 20:33:00 2012
From: nginx-forum at nginx.us (mfouwaaz)
Date: Sun, 01 Jan 2012 15:33:00 -0500
Subject: How to add a new directory under domain root?
In-Reply-To: <49ee2d5d77a7ba176f1299d858100578.NginxMailingListEnglish@forum.nginx.org>
References: <49ee2d5d77a7ba176f1299d858100578.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <069691cfcda07517cf11740d1fd34f50.NginxMailingListEnglish@forum.nginx.org>

Just for the record, things seem to be back to normal when I change the
folder name from images to img.  That`s how I resolved this issue --
until the nginx or css Gods deem otherwise (I tend to lean toward the
former though).

Best wishes for a New Year filled with Peace, Prosperity and Happiness!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220689,220725#msg-220725


From emmiller at gmail.com  Sun Jan  1 16:51:50 2012
From: emmiller at gmail.com (Evan Miller)
Date: Sun, 1 Jan 2012 16:51:50 +0000 (UTC)
Subject: Updated cross-reference (LXR)
Message-ID: <loom.20120101T174508-576@post.gmane.org>

Hello from an old ghost! To ring in the new year I've updated the 
Nginx LXR for the first time in ages:

http://lxr.evanmiller.org/http/source/

It now hosts code from the 1.0.11 and 1.1.12 releases; I've retired
the ancient 0.6/0.7/0.8 sources. Peruse and enjoy!

Evan


From gelonida at gmail.com  Sun Jan  1 22:43:17 2012
From: gelonida at gmail.com (Gelonida N)
Date: Sun, 01 Jan 2012 23:43:17 +0100
Subject: Verify client certificate, but ignore expiration date
In-Reply-To: <jdce01$sn6$1@dough.gmane.org>
References: <jdce01$sn6$1@dough.gmane.org>
Message-ID: <jdqni6$p2u$1@dough.gmane.org>

Hmm no reaction to this question so far.
Does this mean it is impossible:

On 12/27/2011 01:34 PM, Gelonida wrote:
> I wanted to know whether I can configure nginx to verify client
> certificates and reject them if invalid.
> 
> However I would like to exclude the expiration date from the validation
> step.
> 
> The context is rather simple.
> 
> I have some embedded devices trying to connect to a server. The client
> certificates for some of these devices will expire before 
> they will be returned for maintenance
> Instead of disabling client certificates globally  I would like to 'just' ignore
> the expiration date of a selected list of devices
> 
> Ideally I'd like to just ignore the expiration date of a few given
> certificates, but in my current setup even ignoring all expiration dates
> would be an option untill all devices were updated with new certificats
> 
> Is there any setup allowing this?

Alternatively I'd be willing to change the C source of nginx if this
would help me to solve above mentioned issue.

Thanks for any pointers and suggestions.

P.S. I know, that the 'real' answer would be to just avoid above
situation and renew certificates prior to their expiration. However this
is unfortunately not possible for the already deployed devices.





From rainer at ultra-secure.de  Sun Jan  1 22:48:22 2012
From: rainer at ultra-secure.de (Rainer Duffner)
Date: Sun, 1 Jan 2012 23:48:22 +0100
Subject: Verify client certificate, but ignore expiration date
In-Reply-To: <jdce01$sn6$1@dough.gmane.org>
References: <jdce01$sn6$1@dough.gmane.org>
Message-ID: <C9858020-37B0-4DA0-B08A-FF687E01BB4F@ultra-secure.de>


Am 27.12.2011 um 13:34 schrieb Gelonida:

> I wanted to know whether I can configure nginx to verify client certificates and reject them if invalid.
> 
> However I would like to exclude the expiration date from the validation step.
> 
> The context is rather simple.
> 
> I have some embedded devices trying to connect to a server. The client certificate for these devices expired and for a certain time I will be unable to update them.
> 
> Instead of disabling client certificates I would like to 'just' ignore the expiration date.
> 
> Ideally I'd like to just ignore the expiration date of a few given certificates, but in my current setup even ignoring all expiration dates would be an option.
> 
> Is there any setup allowing this?
> 
> Thanks in advance for any suggestion of how to achieve this.
> 



I would suspect that most (all?) validation is done in the SSL-libraries.

As such, you would probably have modify the openssl-source.

I'm no programmer (sitting in a glass house here), but I'd say if you knew how to do that, you wouldn't have asked the original question anyway?.

Instead of trying to find a "quick fix", I would accelerate the project to update the clients.





From ilan at time4learning.com  Mon Jan  2 02:20:23 2012
From: ilan at time4learning.com (Ilan Berkner)
Date: Sun, 1 Jan 2012 21:20:23 -0500
Subject: Help with rewrite rule
Message-ID: <CAOoWpPo1keL2LLRQApOuiirTo-DK2fXe4GiQPz562hc=yeWieQ@mail.gmail.com>

I'm trying to rewrite any png files referenced through the fancybox sub
directory to a directory off of the images folder as follows:

rewrite ^/fancybox/(.+\.png)$ /images/fancybox/$1 last;

This is not working, probably b/c my regex is off the mark.  Any
suggestions?

Thanks
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120101/2ec64a51/attachment.html>

From ilan at time4learning.com  Mon Jan  2 02:27:34 2012
From: ilan at time4learning.com (Ilan Berkner)
Date: Sun, 1 Jan 2012 21:27:34 -0500
Subject: Fwd: Help with rewrite rule
In-Reply-To: <CAOoWpPo1keL2LLRQApOuiirTo-DK2fXe4GiQPz562hc=yeWieQ@mail.gmail.com>
References: <CAOoWpPo1keL2LLRQApOuiirTo-DK2fXe4GiQPz562hc=yeWieQ@mail.gmail.com>
Message-ID: <CAOoWpPos8wy8ddpYCuAkMfsmW3VQqmV-feXFFYL=jpJZ8CgGYQ@mail.gmail.com>

Solved it with this rule:

    rewrite ^(.*)/fancybox/(.+\.png)$ /images/fancybox/$2 last;


---------- Forwarded message ----------
From: Ilan Berkner <ilan at time4learning.com>
Date: Sun, Jan 1, 2012 at 9:20 PM
Subject: Help with rewrite rule
To: nginx at nginx.org


I'm trying to rewrite any png files referenced through the fancybox sub
directory to a directory off of the images folder as follows:

rewrite ^/fancybox/(.+\.png)$ /images/fancybox/$1 last;

This is not working, probably b/c my regex is off the mark.  Any
suggestions?

Thanks




-- 

Ilan Berkner
Chief Technology Officer
Time4Learning.com

6300 NE 1st Ave., Suite 203
Ft. Lauderdale, FL 33334
(954) 771-0914

<http://www.twitter.com/time4learning> <http://www.facebook.com/Time4Learning>


Time4Learning.com - Online interactive curriculum for home use, PreK-8th
Grade.
Time4Writing.com - Online writing tutorials for high, middle, and
elementary school students.
Time4Learning.net - A forum to chat with parents online about kids,
education, parenting and more.
spellingcity.com - Online vocabulary and spelling activities for teachers,
parents and students.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120101/e5b89e78/attachment-0001.html>

From sahmed1020 at gmail.com  Mon Jan  2 02:52:20 2012
From: sahmed1020 at gmail.com (S Ahmed)
Date: Sun, 1 Jan 2012 21:52:20 -0500
Subject: can nginx have client sent the request to a specific server?
Message-ID: <CAG2rwuNC=r8oYxJCydsep4Q=e4RZZ5v2s743wRUN=PVUUmvWTw@mail.gmail.com>

If I have 3 web servers, and 1 nginx sitting in front of them (public ip
mapped to my nginx server).

My other backend servers are:

www1.example.com
www2.example.com
www3.example.com

My users are upload large files, so I don't want the request going through
my nginx front-end server and then to my wwwX.example.com server.

Could it be possible to have nginx map the request to a specific backend
server and then have the request continue with that specific server?

So a request using a proxy I believe goes like this:

client request => proxy => backend-server and then it goes from
backend-server => proxy => response

But I want it to go:

client request => proxy and then backend-server => response => client

i.e. use nginx simply to map the request to a specific server, and then the
client will continue the request/response with that specific server that is
publicly accessible.

Now this request is a single request/response, it will not continue on.
The client is simply uploading a file or posting a file and that's it.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120101/e72c311a/attachment.html>

From piotr.sikora at frickle.com  Mon Jan  2 04:13:16 2012
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Mon, 2 Jan 2012 05:13:16 +0100
Subject: Forcing a Basic Dialog Box.
In-Reply-To: <4EFD2CB2.60809@styleflare.com>
References: <4EFD2CB2.60809@styleflare.com>
Message-ID: <9B701B7CE0CA4CC1B1E11D3113FD2C5A@Desktop>

Hi,
diff inline.

Please note that you'll also need ngx_headers_more module:
https://github.com/agentzh/headers-more-nginx-module

> location / {
>          auth_basic "Restricted";

-auth_basic "Restricted";

>          root        /var/www/web/;
>          auth_request      /auth;
> 
>          try_files $uri $uri/ @upstream;
>      }
> 
> 
>   location /auth {

-location /auth {
+location = /auth {

>              internal;
> 
>              postgres_escape   $user $remote_user;
>              postgres_escape   $pass $remote_passwd;
> 
>              postgres_pass     database;
>              postgres_query    "SELECT login FROM users WHERE 
> login=$user AND pass=$pass";
>              postgres_rewrite  no_rows 403;

-postgres_rewrite  no_rows 403;
+postgres_rewrite  no_rows 401;
+more_set_headers -s 401 'WWW-Authenticate: Basic realm="Restricted"';

>              postgres_output   none;
>          }

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


From quan.nexthop at gmail.com  Mon Jan  2 04:37:36 2012
From: quan.nexthop at gmail.com (Geoge.Q)
Date: Mon, 2 Jan 2012 12:37:36 +0800
Subject: 502 & 504 error, again
Message-ID: <CAP_76MYUE15Mf-YSzR+ivs--TJFJ0rtK0VG1bR=6jZXAYumhZg@mail.gmail.com>

I know it should be a well-known issue.
I searched and try to find out what on earth it is, but I failed.

Step:
1) nginx works as reverse proxy and the version is V1.0.5,
   OS: Debian 5.0.3, Kernel Version 2.6.32

2) Enabled http 1.1,   504 error  is popup in client;
the error log indicates:
-------------------------------------------------
2011/12/30 09:40:06 [error] 11013#0: *1 upstream timed out (110: Connection
timed out) while reading response header from upstream, client:
10.208.127.142, server: 1.1.1.1, request: "POST
/QueryStatManage/StatStudentAction.aspx HTTP/1.1", upstream: "
http://1.1.1.1:80/QueryStatManage/StatStudentAction.aspx", host: "1.1.1.1",
referrer: "http://1.1.1.1/QueryStatManage/StatStudentAction.aspx"
-------------------------------------------------


3) increase the vaule
     > proxy_read_timeout 600s
     > proxy_send_timeout 600s

   this time, it is poped up 502, the error logs indicates
------------------------------------
2011/12/30 09:53:54 [error] 11173#0: *1 upstream prematurely closed
connection while reading response header from upstream, client:
10.208.127.142, server: 1.1.1.1, request: "POST
/QueryStatManage/StatStudentAction.aspx HTTP/1.1", upstream: "
http://1.1.1.1:80/QueryStatManage/StatStudentAction.aspx", host: "1.1.1.1",
referrer: "http://1.1.1.1/QueryStatManage/StatStudentAction.aspx"

4) Someone told me that there is a kernel bug in linux, I am not sure, can
someone see the same issue ?


thanks
George
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120102/abf75360/attachment.html>

From nam.nh at nd24.net  Mon Jan  2 05:10:59 2012
From: nam.nh at nd24.net (Nguyen Hai Nam)
Date: Mon, 2 Jan 2012 12:10:59 +0700
Subject: Reverse proxy for multiple websites
Message-ID: <CAA44a75bN8tKHSzP9eVOTwukBRPJMVLcvAq4iqO91e=rUUwVrw@mail.gmail.com>

Hi,

I'm sticking on configure a reverse proxy, my scenario looks like:

Internet
.......|
.......|__Reverse proxy (listen port 80)
......................|__web1.com (listen port 8080)
......................|__web2.net (listen port 8080)
......................|__web-n.org (listen port 8080)

Assume All website shared one public IP only and each website runs on
independent VPS.

If I create multi server{} block to use proxy_pass directive to web
server inside, it's only web1.com can be reachable (perhaps it's on
top config).

It's highly appreciate if someone advice me ideas to make this working.

Thanks,

-- 
Best regards,
Hai Nam, Nguyen


From jkoh5 at hanmail.net  Mon Jan  2 09:13:19 2012
From: jkoh5 at hanmail.net (=?utf-8?B?7ZWc6re466Oo?=)
Date: Mon, 02 Jan 2012 18:13:19 +0900 (KST)
Subject: anyone can help?
Message-ID: <20120102181319.HM.00000000001b3yY@jkoh5.wwl1576.hanmail.net>

An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120102/d4c102c5/attachment.html>

From david at styleflare.com  Mon Jan  2 09:47:20 2012
From: david at styleflare.com (David)
Date: Mon, 02 Jan 2012 04:47:20 -0500
Subject: Forcing a Basic Dialog Box.
In-Reply-To: <9B701B7CE0CA4CC1B1E11D3113FD2C5A@Desktop>
References: <4EFD2CB2.60809@styleflare.com>
	<9B701B7CE0CA4CC1B1E11D3113FD2C5A@Desktop>
Message-ID: <4F017D28.606@styleflare.com>

Fantastic.

That worked great. I didnt realize I needed the extra $more_headers package.

Appreciate your help.

Thank you!



On 1/1/12 11:13 PM, Piotr Sikora wrote:
> Hi,
> diff inline.
>
> Please note that you'll also need ngx_headers_more module:
> https://github.com/agentzh/headers-more-nginx-module
>
>> location / {
>>          auth_basic "Restricted";
>
> -auth_basic "Restricted";
>
>>          root        /var/www/web/;
>>          auth_request      /auth;
>>
>>          try_files $uri $uri/ @upstream;
>>      }
>>
>>
>>   location /auth {
>
> -location /auth {
> +location = /auth {
>
>>              internal;
>>
>>              postgres_escape   $user $remote_user;
>>              postgres_escape   $pass $remote_passwd;
>>
>>              postgres_pass     database;
>>              postgres_query    "SELECT login FROM users WHERE 
>> login=$user AND pass=$pass";
>>              postgres_rewrite  no_rows 403;
>
> -postgres_rewrite  no_rows 403;
> +postgres_rewrite  no_rows 401;
> +more_set_headers -s 401 'WWW-Authenticate: Basic realm="Restricted"';
>
>>              postgres_output   none;
>>          }
>
> Best regards,
> Piotr Sikora < piotr.sikora at frickle.com >
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From chx1975 at gmail.com  Mon Jan  2 15:40:21 2012
From: chx1975 at gmail.com (Karoly Negyesi)
Date: Mon, 2 Jan 2012 16:40:21 +0100
Subject: variable variables?
Message-ID: <CAOfzkkwTzjOrno+CTPrDLzt2KMFOjs-xFBanC_HR+PgiF1ejiA@mail.gmail.com>

Hi,

Nginx provides the value of the cookie "foo" in $cookie_foo . Is it
possible to read "foo" from the URL? In PHP you would use two dollar
signs for this kind of indirection.

Drupal sets the name of session cookie based on the cookie domain. I
am running SSI (reading from Redis via Webdis) and I would like to
read some user specific information which is stored in a Redis hash
named after the sesion name with a key of the value of the session
cookie. So the SSI command would be something include
virtual="/user/SESSnameofthesession". So I need to rewrite that into
something like /redis/SESSnameofthesession/${COOKIE_$SESSnameofthesession}.

Thanks

Karoly Negyesi

Ps. Sorry if i sent two copies of this.


From nginx-forum at nginx.us  Mon Jan  2 17:08:09 2012
From: nginx-forum at nginx.us (wipeout_dude)
Date: Mon, 02 Jan 2012 12:08:09 -0500
Subject: GZIP not compressing .js and .css files..
Message-ID: <e0598c236e0b7365a7645152ee654e5b.NginxMailingListEnglish@forum.nginx.org>

Hi,

My server doesn't appear to be compressing .js and .css files..

I have the following in my /etc/nginx/nginx.conf..

####
        gzip on;
        gzip_disable "msie6";

        gzip_vary on;
        # gzip_proxied any;
        gzip_comp_level 1;
        gzip_buffers 16 8k;
        gzip_http_version 1.1;
        gzip_types text/plain text/css application/json
application/x-javascript text/xml application/xml application/xml+rss
text/javascript;
####

Pagespeed in Chrome is telling me to enable compression and lists all
the .js and .css files..

Have I missed something?

TIA

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220747,220747#msg-220747


From mdounin at mdounin.ru  Mon Jan  2 23:54:39 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 3 Jan 2012 03:54:39 +0400
Subject: 502 & 504 error, again
In-Reply-To: <CAP_76MYUE15Mf-YSzR+ivs--TJFJ0rtK0VG1bR=6jZXAYumhZg@mail.gmail.com>
References: <CAP_76MYUE15Mf-YSzR+ivs--TJFJ0rtK0VG1bR=6jZXAYumhZg@mail.gmail.com>
Message-ID: <20120102235439.GC67687@mdounin.ru>

Hello!

On Mon, Jan 02, 2012 at 12:37:36PM +0800, Geoge.Q wrote:

> I know it should be a well-known issue.
> I searched and try to find out what on earth it is, but I failed.
> 
> Step:
> 1) nginx works as reverse proxy and the version is V1.0.5,
>    OS: Debian 5.0.3, Kernel Version 2.6.32
> 
> 2) Enabled http 1.1,   504 error  is popup in client;
> the error log indicates:
> -------------------------------------------------
> 2011/12/30 09:40:06 [error] 11013#0: *1 upstream timed out (110: Connection
> timed out) while reading response header from upstream, client:
> 10.208.127.142, server: 1.1.1.1, request: "POST
> /QueryStatManage/StatStudentAction.aspx HTTP/1.1", upstream: "
> http://1.1.1.1:80/QueryStatManage/StatStudentAction.aspx", host: "1.1.1.1",
> referrer: "http://1.1.1.1/QueryStatManage/StatStudentAction.aspx"
> -------------------------------------------------

The "upstream" and "server" seems to be the same, so it looks 
like you've created an infinite proxy loop.

Maxim Dounin


From mdounin at mdounin.ru  Mon Jan  2 23:59:05 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 3 Jan 2012 03:59:05 +0400
Subject: Reverse proxy for multiple websites
In-Reply-To: <CAA44a75bN8tKHSzP9eVOTwukBRPJMVLcvAq4iqO91e=rUUwVrw@mail.gmail.com>
References: <CAA44a75bN8tKHSzP9eVOTwukBRPJMVLcvAq4iqO91e=rUUwVrw@mail.gmail.com>
Message-ID: <20120102235905.GD67687@mdounin.ru>

Hello!

On Mon, Jan 02, 2012 at 12:10:59PM +0700, Nguyen Hai Nam wrote:

> Hi,
> 
> I'm sticking on configure a reverse proxy, my scenario looks like:
> 
> Internet
> .......|
> .......|__Reverse proxy (listen port 80)
> ......................|__web1.com (listen port 8080)
> ......................|__web2.net (listen port 8080)
> ......................|__web-n.org (listen port 8080)
> 
> Assume All website shared one public IP only and each website runs on
> independent VPS.
> 
> If I create multi server{} block to use proxy_pass directive to web
> server inside, it's only web1.com can be reachable (perhaps it's on
> top config).
> 
> It's highly appreciate if someone advice me ideas to make this working.

If many backend sites share the same ip, and you refer all of them 
with the same name (or ip) in proxy_pass directive, you probably 
want to use somethine like

    proxy_set_header Host <name>;

to make sure backend will be able to distinguish them.  Or just

    proxy_set_header Host $host;

to pass what you've got from client.

Maxim Dounin


From mdounin at mdounin.ru  Tue Jan  3 00:33:06 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 3 Jan 2012 04:33:06 +0400
Subject: anyone can help?
In-Reply-To: <20120102181319.HM.00000000001b3yY@jkoh5.wwl1576.hanmail.net>
References: <20120102181319.HM.00000000001b3yY@jkoh5.wwl1576.hanmail.net>
Message-ID: <20120103003305.GE67687@mdounin.ru>

Hello!

On Mon, Jan 02, 2012 at 06:13:19PM +0900, ??? wrote:

>                                    Hello.
> 
> 
>     we have a big problem when we test uploading big file to nginx proxy
>                                   server.
> 
> 
>     In stress test we uploaded only one file with 100M file then the cpu
>                           usage is go up to 95% !!
> 
> 
>                              How can we fix it?

First of all, it's not clear how you managed to get it.  And what 
exactly you get.

You are see nginx using 95% CPU with one 100M POST request, right?  
Could you please provide config, nginx -V output and top(1) snapshot?

Maxim Dounin

p.s. Please post in plain text here, not html.  Thank you.


From agentzh at gmail.com  Tue Jan  3 05:19:19 2012
From: agentzh at gmail.com (agentzh)
Date: Tue, 3 Jan 2012 13:19:19 +0800
Subject: variable variables?
In-Reply-To: <CAOfzkkwTzjOrno+CTPrDLzt2KMFOjs-xFBanC_HR+PgiF1ejiA@mail.gmail.com>
References: <CAOfzkkwTzjOrno+CTPrDLzt2KMFOjs-xFBanC_HR+PgiF1ejiA@mail.gmail.com>
Message-ID: <CAB4Tn6NwD1MYhzneEewqaM7s=fo9M1WJYK1WZeUtK7eWHMda3g@mail.gmail.com>

On Mon, Jan 2, 2012 at 11:40 PM, Karoly Negyesi <chx1975 at gmail.com> wrote:
> Hi,
>
> Nginx provides the value of the cookie "foo" in $cookie_foo . Is it
> possible to read "foo" from the URL? In PHP you would use two dollar
> signs for this kind of indirection.
>

Well, it is trivial with ngx_lua's set_by_lua config directive:

    location /blah {
        set_by_lua $arg_value
            'local arg_name = ngx.var.cookie_foo
             return ngx.var["arg_" .. arg_name]';

        # reference $arg_value in later phase config directives here...
    }

See the documentation for details:

    http://wiki.nginx.org/HttpLuaModule#set_by_lua

Regards,
-agentzh


From agentzh at gmail.com  Tue Jan  3 05:25:40 2012
From: agentzh at gmail.com (agentzh)
Date: Tue, 3 Jan 2012 13:25:40 +0800
Subject: variable variables?
In-Reply-To: <CAOfzkkwTzjOrno+CTPrDLzt2KMFOjs-xFBanC_HR+PgiF1ejiA@mail.gmail.com>
References: <CAOfzkkwTzjOrno+CTPrDLzt2KMFOjs-xFBanC_HR+PgiF1ejiA@mail.gmail.com>
Message-ID: <CAB4Tn6N4eH8cPy0G=6cgEk66KYeGqgOy-Bvcz4jw1aum23giqA@mail.gmail.com>

On Mon, Jan 2, 2012 at 11:40 PM, Karoly Negyesi <chx1975 at gmail.com> wrote:
>
> Drupal sets the name of session cookie based on the cookie domain. I
> am running SSI (reading from Redis via Webdis) and I would like to
> read some user specific information which is stored in a Redis hash
> named after the sesion name with a key of the value of the session
> cookie. So the SSI command would be something include
> virtual="/user/SESSnameofthesession". So I need to rewrite that into
> something like /redis/SESSnameofthesession/${COOKIE_$SESSnameofthesession}.
>

I think you may consider replacing webdis with ngx_redis2:
http://wiki.nginx.org/HttpRedis2Module

Also, you may consider replacing SSI with content_by_lua because the
latter provides a full-fledged scripting language as well as the same
nonblocking network I/O capabilities (with ngx_redis2, among other
things).

Regards,
-agentzh


From quan.nexthop at gmail.com  Tue Jan  3 12:37:34 2012
From: quan.nexthop at gmail.com (Geoge.Q)
Date: Tue, 3 Jan 2012 20:37:34 +0800
Subject: 502 & 504 error, again
In-Reply-To: <20120102235439.GC67687@mdounin.ru>
References: <CAP_76MYUE15Mf-YSzR+ivs--TJFJ0rtK0VG1bR=6jZXAYumhZg@mail.gmail.com>
	<20120102235439.GC67687@mdounin.ru>
Message-ID: <CAP_76MZJOCiPr_eVSfbuA=0Xv99EmmrGEYQbiFSAMD5qRBi50w@mail.gmail.com>

thanks Max. I will check and update you.

George

On Tue, Jan 3, 2012 at 7:54 AM, Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
>
> On Mon, Jan 02, 2012 at 12:37:36PM +0800, Geoge.Q wrote:
>
> > I know it should be a well-known issue.
> > I searched and try to find out what on earth it is, but I failed.
> >
> > Step:
> > 1) nginx works as reverse proxy and the version is V1.0.5,
> >    OS: Debian 5.0.3, Kernel Version 2.6.32
> >
> > 2) Enabled http 1.1,   504 error  is popup in client;
> > the error log indicates:
> > -------------------------------------------------
> > 2011/12/30 09:40:06 [error] 11013#0: *1 upstream timed out (110:
> Connection
> > timed out) while reading response header from upstream, client:
> > 10.208.127.142, server: 1.1.1.1, request: "POST
> > /QueryStatManage/StatStudentAction.aspx HTTP/1.1", upstream: "
> > http://1.1.1.1:80/QueryStatManage/StatStudentAction.aspx", host:
> "1.1.1.1",
> > referrer: "http://1.1.1.1/QueryStatManage/StatStudentAction.aspx"
> > -------------------------------------------------
>
> The "upstream" and "server" seems to be the same, so it looks
> like you've created an infinite proxy loop.
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120103/baf0426a/attachment.html>

From cabbar at gmail.com  Tue Jan  3 13:32:36 2012
From: cabbar at gmail.com (Cabbar Duzayak)
Date: Tue, 3 Jan 2012 15:32:36 +0200
Subject: Empty /error log lines (a lot of them)
Message-ID: <CA+w7OpK_uNxMz4-1dfCL25HwmmjDe0HyfmV3OBVfvOGZFDrP5g@mail.gmail.com>

Hi,

I am using NGINX 0.8.54 on Ubuntu 11.04, and I am getting a lot of
weird log entries in nginx access file. The structure is somewhat like
below, but as  you can see there is no method, URL, referer, user
agent, etc. And, I see bunch of entries for different IP addresses (3
to 30-40 entries per IP)

It might be a browser issue, but since I am seeing a lot of these, I
just wanted to ask here if anyone has seen this before.

Thanks...


<USER_IP_ADDRESS> - - [07/Oct/2011:17:11:22 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:11:22 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:17:52 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:17:52 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:17:52 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:17:52 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:20:02 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:20:02 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:20:02 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:46:56 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:46:56 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:46:56 +0300] "-" 400 0 "-" "-"
<USER_IP_ADDRESS> - - [07/Oct/2011:17:49:46 +0300] "-" 400 0 "-" "-"


From stef at scaleengine.com  Mon Jan  2 11:23:47 2012
From: stef at scaleengine.com (Stefan Caunter)
Date: Mon, 2 Jan 2012 06:23:47 -0500
Subject: anyone can help?
In-Reply-To: <20120102181319.HM.00000000001b3yY@jkoh5.wwl1576.hanmail.net>
References: <20120102181319.HM.00000000001b3yY@jkoh5.wwl1576.hanmail.net>
Message-ID: <CALdpnrwq3OqOf5eEu_-XjxikBmfi8T=L0c2WvK_fARjxDiwx_A@mail.gmail.com>

On Mon, Jan 2, 2012 at 4:13 AM, ??? <jkoh5 at hanmail.net> wrote:
>
> Hello.
>
>
>
> we have a big problem?when we test?uploading big file to nginx proxy server.
>
>
>
> In stress test we uploaded only one file with 100M file then the cpu usage is go up to 95% !!
>
>
>
> How can we fix it?
>

Please provide relevant logging.

What is the hardware profile and OS?

What is the nginx configuration?

What is the application?

----

Stefan Caunter


From stef at scaleengine.com  Mon Jan  2 11:33:29 2012
From: stef at scaleengine.com (Stefan Caunter)
Date: Mon, 2 Jan 2012 06:33:29 -0500
Subject: 502 & 504 error, again
In-Reply-To: <CAP_76MYUE15Mf-YSzR+ivs--TJFJ0rtK0VG1bR=6jZXAYumhZg@mail.gmail.com>
References: <CAP_76MYUE15Mf-YSzR+ivs--TJFJ0rtK0VG1bR=6jZXAYumhZg@mail.gmail.com>
Message-ID: <CALdpnryaJo4wMnW8Oa1xek-yCORxUY5TsFqFmgoFu6kKD1Zg_g@mail.gmail.com>

On Sun, Jan 1, 2012 at 11:37 PM, Geoge.Q <quan.nexthop at gmail.com> wrote:
> I know it should be a well-known issue.
> I searched and try to find out what on earth it is, but I failed.
>
> Step:
> 1) nginx works as reverse proxy and the version is V1.0.5,
> ?? OS: Debian 5.0.3, Kernel Version 2.6.32
>
> 2) Enabled http 1.1,?? 504 error? is popup in client;

You enabled the HTTP/1.1 protocol? Where? What were you doing before?

> the error log indicates:
> -------------------------------------------------
> 2011/12/30 09:40:06 [error] 11013#0: *1 upstream timed out (110: Connection
> timed out) while reading response header from upstream, client:
> 10.208.127.142, server: 1.1.1.1, request: "POST
> /QueryStatManage/StatStudentAction.aspx HTTP/1.1", upstream:
> "http://1.1.1.1:80/QueryStatManage/StatStudentAction.aspx", host: "1.1.1.1",
> referrer: "http://1.1.1.1/QueryStatManage/StatStudentAction.aspx"
> -------------------------------------------------
>

How large is the POST?

>
> 3) increase the vaule
> ???? > proxy_read_timeout 600s
> ???? > proxy_send_timeout 600s
>
> ?? this time, it is poped up 502, the error logs indicates
> ------------------------------------
> 2011/12/30 09:53:54 [error] 11173#0: *1 upstream prematurely closed
> connection while reading response header from upstream, client:
> 10.208.127.142, server: 1.1.1.1, request: "POST
> /QueryStatManage/StatStudentAction.aspx HTTP/1.1", upstream:
> "http://1.1.1.1:80/QueryStatManage/StatStudentAction.aspx", host: "1.1.1.1",
> referrer: "http://1.1.1.1/QueryStatManage/StatStudentAction.aspx"
>

Does the back end application work without nginx as proxy? What is the
nginx configuration? We need details to help you.

> 4) Someone told me that there is a kernel bug in linux, I am not sure, can
> someone see the same issue ?

Irrelevant. These are HTTP errors, with HTTP logging. If you have
kernel errors, send your logs to the appropriate maintainers :)

You are running an ASP.NET application. You have to correctly set
proxy timeouts and buffer sizes to proxy to ASP applications, assuming
the application works.


Stefan Caunter
http://scaleengine.com/contact


From nginx-forum at nginx.us  Tue Jan  3 16:59:52 2012
From: nginx-forum at nginx.us (fernandokosh)
Date: Tue, 03 Jan 2012 11:59:52 -0500
Subject: How to solve the problem of "405 not allowed"?
In-Reply-To: <63d620952a4a8185eb9a41fd44e226f0.NginxMailingListEnglish@forum.nginx.org>
References: <c1b6567d6676f62cdcd658d71a9151cc.NginxMailingList@forum.nginx.org>
	<0f5a7655c4c4966f79b764d0c974d75a.NginxMailingListEnglish@forum.nginx.org>
	<63d620952a4a8185eb9a41fd44e226f0.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <c6a6aecbaf4eab9af3cbb728e92e3aa4.NginxMailingListEnglish@forum.nginx.org>

I have the same issue here running Rails 2 apps virtual hosts on Debian
Squeeze.
In my case I, the error occur when I perform update with the PUT
method.
That's solve my problem:

location / {
error_page 405 = $uri;
try_files $uri @unicorn_app;
}

Thank you!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,2414,220771#msg-220771


From r at roze.lv  Tue Jan  3 17:18:45 2012
From: r at roze.lv (Reinis Rozitis)
Date: Tue, 3 Jan 2012 19:18:45 +0200
Subject: image_filter on error_page location
Message-ID: <EE68C4EA214845FA95150744903CD0BF@DD21>

Hello,
I can't figure how to solve such issue - nginx seems not to apply image_filter in case the picture comes from upstream via internal 
error_page.


I have something like (if picture is prefixed with sm_ a dynamic thumbnail is made and the original saved locally):

location ~ /sm_ {
        rewrite (.*)/sm_(.*) $1/$2 break;
        image_filter   crop  100 100;
        error_page      404 = @store;
}

location @store {
        internal;
        proxy_pass           http://backend/$uri;
        proxy_store          on;

}


If the file doesn't exist on the server client gets the original/full sized image, on the second request when the file is stored on 
the nginx box the image/response gets resized.


Also requesting directly from the upstream within the location block:

location ~ /sm_ {
        rewrite (.*)/sm_(.*) $1/$2 break;
        image_filter   crop  100 100;
        proxy_pass           http://backend/$uri;
}

works as expected but in this case obviously no local copy (which is necessary) is made.

Any suggestions?


wbr
rr 


From mdounin at mdounin.ru  Tue Jan  3 22:00:29 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Wed, 4 Jan 2012 02:00:29 +0400
Subject: Empty /error log lines (a lot of them)
In-Reply-To: <CA+w7OpK_uNxMz4-1dfCL25HwmmjDe0HyfmV3OBVfvOGZFDrP5g@mail.gmail.com>
References: <CA+w7OpK_uNxMz4-1dfCL25HwmmjDe0HyfmV3OBVfvOGZFDrP5g@mail.gmail.com>
Message-ID: <20120103220029.GJ67687@mdounin.ru>

Hello!

On Tue, Jan 03, 2012 at 03:32:36PM +0200, Cabbar Duzayak wrote:

> Hi,
> 
> I am using NGINX 0.8.54 on Ubuntu 11.04, and I am getting a lot of
> weird log entries in nginx access file. The structure is somewhat like
> below, but as  you can see there is no method, URL, referer, user
> agent, etc. And, I see bunch of entries for different IP addresses (3
> to 30-40 entries per IP)
> 
> It might be a browser issue, but since I am seeing a lot of these, I
> just wanted to ask here if anyone has seen this before.
> 
> Thanks...
> 
> 
> <USER_IP_ADDRESS> - - [07/Oct/2011:17:11:22 +0300] "-" 400 0 "-" "-"

[...]

Such lines appear when client opens a connection and doesn't send 
valid request (e.g. just closes the connection).

Some modern browsers (e.g. Chrome) tend to open several extra 
connections "just in case" and close them once they realize they 
have no requests to send (and thus they cause such lines to 
appear).

Another common reason is health checks from external balancers, 
but it's probably not your case.

Maxim Dounin


From mdounin at mdounin.ru  Tue Jan  3 22:08:19 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Wed, 4 Jan 2012 02:08:19 +0400
Subject: image_filter on error_page location
In-Reply-To: <EE68C4EA214845FA95150744903CD0BF@DD21>
References: <EE68C4EA214845FA95150744903CD0BF@DD21>
Message-ID: <20120103220819.GK67687@mdounin.ru>

Hello!

On Tue, Jan 03, 2012 at 07:18:45PM +0200, Reinis Rozitis wrote:

> Hello,
> I can't figure how to solve such issue - nginx seems not to apply
> image_filter in case the picture comes from upstream via internal
> error_page.
> 
> 
> I have something like (if picture is prefixed with sm_ a dynamic thumbnail is made and the original saved locally):
> 
> location ~ /sm_ {
>        rewrite (.*)/sm_(.*) $1/$2 break;
>        image_filter   crop  100 100;
>        error_page      404 = @store;
> }
> 
> location @store {
>        internal;
>        proxy_pass           http://backend/$uri;
>        proxy_store          on;
> 
> }
> 
> 
> If the file doesn't exist on the server client gets the
> original/full sized image, on the second request when the file is
> stored on the nginx box the image/response gets resized.

This is expected behaviour: you have no image_filter configured in 
location @store, which handles the request if file doesn't exists.

Obvious fix is to add image_filter to location @store, i.e.

    location @store {
        internal;
        proxy_pass    http://backend/$uri;
        proxy_store   on;
        image_filter  crop  100 100;
    }

Maxim Dounin


From nginx-forum at nginx.us  Tue Jan  3 22:35:31 2012
From: nginx-forum at nginx.us (ronjpark)
Date: Tue, 03 Jan 2012 17:35:31 -0500
Subject: Optimize nginx for uploading/download huge files..
Message-ID: <aa3060b95f2e21e88b0942423f46407b.NginxMailingListEnglish@forum.nginx.org>

Hello everyone,

Firstly, I'm sorry for posting the same question here and there (How to
forum).
It is urgent issue to me.

Could you give me a tip for which directives and values I have to set
for the optimization?

I installed nginx on virtual machine as reverse proxy server.
It checks the authentication and bypasses the client request to the
backend servers.
The most transactions are to upload/download files, from 1M to 1G...

The CPU usage hits 100% when the 1M file download request reaches around
25 TPS..
around 20 TPS for upload request..

I set client_body_buffer_size, proxy_buffer_size, proxy_buffers and so
on..
But not so good yet.. :(

client_body_buffer_size is set to 2M and it helps to increase TPS for
uploading, but not enough.
proxy_buffer_size = 1M and proxy_buffers = 4 1M help to increase TPS for
downloading little bit, but not so much..
Other directives are set by its default.

The below is the result of profiling with google_proftools_module.
As you can see, the most time is spent on readv() and writev().

Please give me any advice or your experience.
Thank you for your time.

Total: 22244 samples
11928 53.6% 53.6% 11930 53.6% readv
8212 36.9% 90.5% 8212 36.9% writev
907 4.1% 94.6% 907 4.1% recv
359 1.6% 96.2% 359 1.6% __connect_nocancel
248 1.1% 97.3% 248 1.1% __close_nocancel
35 0.2% 97.5% 46 0.2% _IO_str_pbackfail
34 0.2% 97.7% 34 0.2% __write_nocancel
31 0.1% 97.8% 31 0.1% socket
28 0.1% 97.9% 28 0.1% memset
and
Total: 64475 samples
34117 52.9% 52.9% 34117 52.9% writev
14173 22.0% 74.9% 14181 22.0% readv
11981 18.6% 93.5% 11981 18.6% recv
1046 1.6% 95.1% 1049 1.6% __connect_nocancel
878 1.4% 96.5% 878 1.4% __close_nocancel
343 0.5% 97.0% 343 0.5% epoll_wait
160 0.2% 97.2% 160 0.2% __write_nocancel
89 0.1% 97.4% 89 0.1% __xstat64
81 0.1% 97.5% 115 0.2% _IO_str_pbackfail


Here is the information of my VM:
ronjpark at PSEProxy:~/release/REL.0.13$ cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 2
model name : QEMU Virtual CPU version 0.14.1
stepping : 3
cpu MHz : 2393.998
cache size : 4096 KB
fpu : yes
fpu_exception : yes
cpuid level : 4
wp : yes
flags : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36
clflush mmx fxsr sse sse2 syscall nx lm up rep_good nopl pni cx16 popcnt
hypervisor lahf_lm
bogomips : 4787.99
clflush size : 64
cache_alignment : 64
address sizes : 40 bits physical, 48 bits virtual
power management:

ronjpark at PSEProxy:~/release/REL.0.13$ cat /proc/meminfo
MemTotal: 4057052 kB
MemFree: 2651832 kB
Buffers: 80984 kB
Cached: 1100692 kB
SwapCached: 3856 kB
Active: 325744 kB
Inactive: 910176 kB
Active(anon): 48884 kB
Inactive(anon): 5684 kB
Active(file): 276860 kB
Inactive(file): 904492 kB
Unevictable: 0 kB
Mlocked: 0 kB
SwapTotal: 2095100 kB
SwapFree: 2082688 kB
Dirty: 4 kB
Writeback: 0 kB
AnonPages: 51356 kB
Mapped: 15740 kB
Shmem: 324 kB
Slab: 66252 kB
SReclaimable: 56796 kB
SUnreclaim: 9456 kB
KernelStack: 1160 kB
PageTables: 4124 kB
NFS_Unstable: 0 kB
Bounce: 0 kB
WritebackTmp: 0 kB
CommitLimit: 4123624 kB
Committed_AS: 176112 kB
VmallocTotal: 34359738367 kB
VmallocUsed: 28816 kB
VmallocChunk: 34359707236 kB
HardwareCorrupted: 0 kB
HugePages_Total: 0
HugePages_Free: 0
HugePages_Rsvd: 0
HugePages_Surp: 0
Hugepagesize: 2048 kB
DirectMap4k: 42996 kB
DirectMap2M: 4151296 kB
ronjpark at PSEProxy:~/release/REL.0.13$ uname -a
Linux PSEProxy 2.6.38-8-server #42-Ubuntu SMP Mon Apr 11 03:49:04 UTC
2011 x86_64 x86_64 x86_64 GNU/Linux

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220778,220778#msg-220778


From r at roze.lv  Wed Jan  4 01:27:07 2012
From: r at roze.lv (Reinis Rozitis)
Date: Wed, 4 Jan 2012 03:27:07 +0200
Subject: image_filter on error_page location
In-Reply-To: <20120103220819.GK67687@mdounin.ru>
References: <EE68C4EA214845FA95150744903CD0BF@DD21>
	<20120103220819.GK67687@mdounin.ru>
Message-ID: <DE72ED430A2F4B1C9254ECAA480E5B26@NeiRoze>

> This is expected behaviour: you have no image_filter configured in 
> location @store, which handles the request if file doesn't exists.
> Obvious fix is to add image_filter to location @store, i.e.

I see.
Got a feeling that will have to define a seperate 404 location for each 
thumbnail block (since there are bunch and all have different image_filter 
params).

rr 


From m.kliewe at team.mail.de  Wed Jan  4 01:55:57 2012
From: m.kliewe at team.mail.de (Michael Kliewe)
Date: Wed, 04 Jan 2012 02:55:57 +0100
Subject: Empty /error log lines (a lot of them)
In-Reply-To: <CA+w7OpK_uNxMz4-1dfCL25HwmmjDe0HyfmV3OBVfvOGZFDrP5g@mail.gmail.com>
References: <CA+w7OpK_uNxMz4-1dfCL25HwmmjDe0HyfmV3OBVfvOGZFDrP5g@mail.gmail.com>
Message-ID: <4F03B1AD.4090207@team.mail.de>

Hi,

I also have a lot of these lines in my logs on all nginx webservers, 
they are a result of loadbalancers or nagios/icinga or something that is 
checking your webservers port for availability, some kind of half open 
connection without a full http request. If there would be any 
configuration setting to disable these entries that would be great, 
would save I/O and disk space ;-)

Michael

Am 03.01.2012 14:32, schrieb Cabbar Duzayak:
> Hi,
>
> I am using NGINX 0.8.54 on Ubuntu 11.04, and I am getting a lot of
> weird log entries in nginx access file. The structure is somewhat like
> below, but as  you can see there is no method, URL, referer, user
> agent, etc. And, I see bunch of entries for different IP addresses (3
> to 30-40 entries per IP)
>
> It might be a browser issue, but since I am seeing a lot of these, I
> just wanted to ask here if anyone has seen this before.
>
> Thanks...
>
>
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:11:22 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:11:22 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:17:52 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:17:52 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:17:52 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:17:52 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:20:02 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:20:02 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:20:02 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:46:56 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:46:56 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:46:56 +0300] "-" 400 0 "-" "-"
> <USER_IP_ADDRESS>  - - [07/Oct/2011:17:49:46 +0300] "-" 400 0 "-" "-"
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From nbubingo at gmail.com  Wed Jan  4 04:43:45 2012
From: nbubingo at gmail.com (=?GB2312?B?0qbOsLHz?=)
Date: Wed, 4 Jan 2012 12:43:45 +0800
Subject: Updated cross-reference (LXR)
In-Reply-To: <loom.20120101T174508-576@post.gmane.org>
References: <loom.20120101T174508-576@post.gmane.org>
Message-ID: <CAGxuLK1BPoBO9n4emLoiaf1SOm3Sa-hdQgCd-Y5fAUQNub2isQ@mail.gmail.com>

Thanks.  Evan.

2012/1/2 Evan Miller <emmiller at gmail.com>:
> Hello from an old ghost! To ring in the new year I've updated the
> Nginx LXR for the first time in ages:
>
> http://lxr.evanmiller.org/http/source/
>
> It now hosts code from the 1.0.11 and 1.1.12 releases; I've retired
> the ancient 0.6/0.7/0.8 sources. Peruse and enjoy!
>
> Evan
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From cabbar at gmail.com  Wed Jan  4 07:03:51 2012
From: cabbar at gmail.com (Cabbar Duzayak)
Date: Wed, 4 Jan 2012 09:03:51 +0200
Subject: Empty /error log lines (a lot of them)
In-Reply-To: <4F03B1AD.4090207@team.mail.de>
References: <CA+w7OpK_uNxMz4-1dfCL25HwmmjDe0HyfmV3OBVfvOGZFDrP5g@mail.gmail.com>
	<4F03B1AD.4090207@team.mail.de>
Message-ID: <CA+w7OpL2C1cbPV-KpQhTuJHPMTkYDC_+AOD+_7cVSjs=CH_1_g@mail.gmail.com>

Hi,

Upon further investigation, it turned out to be coming from chrome
browser actually... I could not find why it is doing that (it was not
shown in its network details page nor developer tools), but every once
in a while it is going crazy and sending 4-5 requests at a time. And,
it looked like these were generated from the browser right after a
facebook call. And number of requests varies as well, sometimes it is
just one request, sometimes it is 4 at a time.

BTW, FF is not doing this.

As Michael was saying, this is causing unnecessary I/O, disk space,
etc. And, wish there was a way of disabling this!

Can we at least tell Nginx not to log these? Something like, if there
is no http method / url, skip the log?

Thanks...


On Wed, Jan 4, 2012 at 3:55 AM, Michael Kliewe <m.kliewe at team.mail.de> wrote:
> Hi,
>
> I also have a lot of these lines in my logs on all nginx webservers, they
> are a result of loadbalancers or nagios/icinga or something that is checking
> your webservers port for availability, some kind of half open connection
> without a full http request. If there would be any configuration setting to
> disable these entries that would be great, would save I/O and disk space ;-)
>
> Michael
>
> Am 03.01.2012 14:32, schrieb Cabbar Duzayak:
>>
>> Hi,
>>
>> I am using NGINX 0.8.54 on Ubuntu 11.04, and I am getting a lot of
>> weird log entries in nginx access file. The structure is somewhat like
>> below, but as ?you can see there is no method, URL, referer, user
>> agent, etc. And, I see bunch of entries for different IP addresses (3
>> to 30-40 entries per IP)
>>
>> It might be a browser issue, but since I am seeing a lot of these, I
>> just wanted to ask here if anyone has seen this before.
>>
>> Thanks...
>>
>>
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:11:22 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:11:22 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:11:52 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:12:32 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:13:22 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:17:52 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:17:52 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:17:52 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:17:52 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:20:02 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:20:02 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:20:02 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:45:56 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:46:16 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:46:56 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:46:56 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:46:56 +0300] "-" 400 0 "-" "-"
>> <USER_IP_ADDRESS> ?- - [07/Oct/2011:17:49:46 +0300] "-" 400 0 "-" "-"
>>
>> _______________________________________________
>> nginx mailing list
>> nginx at nginx.org
>> http://mailman.nginx.org/mailman/listinfo/nginx
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From agentzh at gmail.com  Wed Jan  4 11:48:35 2012
From: agentzh at gmail.com (agentzh)
Date: Wed, 4 Jan 2012 19:48:35 +0800
Subject: Is nginx vulnerable to the Hash Table Vulnerability (n.runs AG)?
In-Reply-To: <CAB4Tn6PT9ZdEwt3eKzfXYS=wyT-A120RH7R_QgDrs9=a9DVi_A@mail.gmail.com>
References: <CAEHYtVjfNoA6uYQE+DmOFEzj05znaMoNmUGRa6FtOTsD2EOJQQ@mail.gmail.com>
	<CAB4Tn6OU5h3m5GRkZUTynoixJ9MedyXvduhXh4kCNQvkzrc3TQ@mail.gmail.com>
	<1285211291852136747@unknownmsgid>
	<CAB4Tn6PT9ZdEwt3eKzfXYS=wyT-A120RH7R_QgDrs9=a9DVi_A@mail.gmail.com>
Message-ID: <CAB4Tn6Nb976BK7kvoBW57bysgM4QN3BHGysFny5bh_CZRvBGmg@mail.gmail.com>

On Sun, Jan 1, 2012 at 10:20 PM, agentzh <agentzh at gmail.com> wrote:
> The current released versions of ngx_lua does have this vulnerability
> in its ngx.req.get_uri_args() and ngx.req.get_post_args() functions.
> I've already worked out a patch for these two functions in ngx_lua's
> git max-args branch here:
>
> ? ?https://github.com/chaoslawful/lua-nginx-module/commit/75876
>
> With this patch, both of these functions will only parse 100 query
> args at most. And one can specify a custom maximum number of args
> parsed with an optional function argument (default to 100) and
> enforcing unlimited parsing by specifying a zero number.
>
> This patch (as well as this branch) will be merged into the master
> branch in 3 Jan.
>

I've also added similar protections to ngx.req.get_headers():

    http://wiki.nginx.org/HttpLuaModule#ngx.req.get_headers

All of these changes have been released as ngx_lua 0.3.1rc45:

   https://github.com/chaoslawful/lua-nginx-module/tags

and also included in the ngx_openresty bundle's devel version 1.0.10.39:

   http://openresty.org/#Download

Feedback welcome!

Best,
-agentzh


From nginx-forum at nginx.us  Wed Jan  4 14:12:12 2012
From: nginx-forum at nginx.us (od3n)
Date: Wed, 04 Jan 2012 09:12:12 -0500
Subject: rewrite rule custom.php to custom.css
Message-ID: <448df5c84ba71ba759f9506d759831e1.NginxMailingListEnglish@forum.nginx.org>

hi,

how can i rename custom.php to custom.css for my wordpress custom css.

let say i put my custom.php inside my theme folder

http://example.com/wp-content/themes/mytheme/assets/css/custom.php

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220792,220792#msg-220792


From appa at perusio.net  Wed Jan  4 14:28:07 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Wed, 04 Jan 2012 14:28:07 +0000
Subject: rewrite rule custom.php to custom.css
In-Reply-To: <448df5c84ba71ba759f9506d759831e1.NginxMailingListEnglish@forum.nginx.org>
References: <448df5c84ba71ba759f9506d759831e1.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <878vln8ruw.wl%appa@perusio.net>

On 4 Jan 2012 14h12 WET, nginx-forum at nginx.us wrote:

> hi,
>
> how can i rename custom.php to custom.css for my wordpress custom
> css.

You mean rewrite or redirect I suppose.

> let say i put my custom.php inside my theme folder
>
> http://example.com/wp-content/themes/mytheme/assets/css/custom.php

Try:

location ^~ /wp-content/themes/mytheme/assets/css/custom.php {
         return 302 /wp-content/themes/mytheme/assets/css/custom.css;
}

--- appa


From nginx-forum at nginx.us  Wed Jan  4 14:41:26 2012
From: nginx-forum at nginx.us (od3n)
Date: Wed, 04 Jan 2012 09:41:26 -0500
Subject: rewrite rule custom.php to custom.css
In-Reply-To: <878vln8ruw.wl%appa@perusio.net>
References: <878vln8ruw.wl%appa@perusio.net>
Message-ID: <cc3994bc60fdf365c3d1b28d38937b12.NginxMailingListEnglish@forum.nginx.org>

it's not redirect. i have .php file, but no .css file.

if in apache, rewrite rule is like this :

<IfModule mod_rewrite.c>
	RewriteEngine On
	RewriteRule custom.css custom.php [L,QSA]
</IfModule>

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220792,220794#msg-220794


From appa at perusio.net  Wed Jan  4 14:53:01 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Wed, 04 Jan 2012 14:53:01 +0000
Subject: rewrite rule custom.php to custom.css
In-Reply-To: <cc3994bc60fdf365c3d1b28d38937b12.NginxMailingListEnglish@forum.nginx.org>
References: <878vln8ruw.wl%appa@perusio.net>
	<cc3994bc60fdf365c3d1b28d38937b12.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <877h178qpe.wl%appa@perusio.net>

On 4 Jan 2012 14h41 WET, nginx-forum at nginx.us wrote:

> it's not redirect. i have .php file, but no .css file.
>
> if in apache, rewrite rule is like this :
>
> <IfModule mod_rewrite.c>
> 	RewriteEngine On
> 	RewriteRule custom.css custom.php [L,QSA]
> </IfModule>
>

It's the other way around:

location ^~ /wp-content/themes/mytheme/assets/css/custom.css {
    return 302 /wp-content/themes/mytheme/assets/css/custom.php;
}

--- appa


From moseleymark at gmail.com  Wed Jan  4 17:29:20 2012
From: moseleymark at gmail.com (Mark Moseley)
Date: Wed, 4 Jan 2012 09:29:20 -0800
Subject: Netcraft Survey
Message-ID: <CAOH1cHnR+F=LkAr=kVsfZc=ZJe6zyC17REi_VnSceu=+f1xf8Q@mail.gmail.com>

Congrats on passing M$ in the Active Site category. We've been using
nginx for a few years now, to front Apache servers, with dramatic
results, so I'm very grateful for all your efforts and think the rise
in nginx popularity is well-deserved.


From mit at stagename.com  Wed Jan  4 18:27:34 2012
From: mit at stagename.com (Mit Rowe)
Date: Wed, 4 Jan 2012 13:27:34 -0500
Subject: Empty /error log lines (a lot of them)
In-Reply-To: <CA+w7OpL2C1cbPV-KpQhTuJHPMTkYDC_+AOD+_7cVSjs=CH_1_g@mail.gmail.com>
References: <CA+w7OpK_uNxMz4-1dfCL25HwmmjDe0HyfmV3OBVfvOGZFDrP5g@mail.gmail.com>
	<4F03B1AD.4090207@team.mail.de>
	<CA+w7OpL2C1cbPV-KpQhTuJHPMTkYDC_+AOD+_7cVSjs=CH_1_g@mail.gmail.com>
Message-ID: <CAFWPQt6vuXEp7hWg_t=qqEAN0yEcGFa-o0QjJz82Dj0womAOJA@mail.gmail.com>

On Wed, Jan 4, 2012 at 2:03 AM, Cabbar Duzayak <cabbar at gmail.com> wrote:

> Hi,
>
>
[..snip..]


> Can we at least tell Nginx not to log these? Something like, if there
> is no http method / url, skip the log?
>
>

It's very likely that there's no Host header sent, so it's being directed
to the server set as "default".

In my setup i define a server{} section for every real host, and then add a
separate "catch all" one for everything else and turn off the logs when not
debugging.
If you are able to do something similar, it should capture those spurious
connections, direct them to the default host, and not log the lines

#catch all
    server {
        listen 80 default_server;
        server_name localhost.domain.com localhost 127.0.0.1 xx.xx.xx.xx;
#where xx.xx.xx.xx is the local IP
        access_log off;
        location / {
            #local
            allow 127.0.0.1;
            #office router
            allow xx.xx.xx.xx;
            #protect
            deny all;
        }
    }

#real server(s)
   server {
      access_log /var/log/nginx/access.log;
      listen 80;
      server_name .domain.com; #matches www.domain.com and domain.com
     location / {
        #[..snip..]
     }
   }




-- 
Will 'Mit' Rowe
Stagename*
*1-866-326-3098
mit at stagename.com <josh at stagename.com>
www.stagename.com
Twitter: @stagename

*The information transmitted is intended only for the person or entity to
which it is addressed and may contain confidential and/or privileged
material. Any review, retransmission, dissemination or other use of this
information by persons or entities other than the intended recipient is
prohibited. If you received this transmission in error, please contact the
sender and delete all material contained herein from your computer.*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120104/5de8c352/attachment.html>

From nginx-forum at nginx.us  Wed Jan  4 19:05:59 2012
From: nginx-forum at nginx.us (fxofxo)
Date: Wed, 04 Jan 2012 14:05:59 -0500
Subject: Simply change the listen port ,does not work
Message-ID: <a88e1717bd60a8d0c468ddf8dccb680c.NginxMailingListEnglish@forum.nginx.org>

Hi, I just installed nginx-1.0.11 , and start it by default , view
http://xxx.xxx.xxx.xxx:80  ,and nginx works
Then I change the listen port in default.conf  to 82,  kill nginx ,
restart , and view http://xxx.xxx.xxx.xxx:82, nothing happened, nginx 
failed. 
Could anyone tell me what's wrong with it ? Thanks very much.

server {
    listen       82;
    server_name  localhost;

    #charset koi8-r;
    #access_log  /var/log/nginx/log/host.access.log  main;

    location / {
        root   /usr/share/nginx/html;
        index  index.html index.htm;
    }

    #error_page  404              /404.html;

    # redirect server error pages to the static page /50x.html
    #
    error_page   500 502 503 504  /50x.html;
    location = /50x.html {
        root   /usr/share/nginx/html;
    }

    # 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$ {
    #    root           html;
    #    fastcgi_pass   127.0.0.1:9000;
    #    fastcgi_index  index.php;
    #    fastcgi_param  SCRIPT_FILENAME  /scripts$fastcgi_script_name;
    #    include        fastcgi_params;
    #}

    # deny access to .htaccess files, if Apache's document root
    # concurs with nginx's one
    #
    #location ~ /\.ht {
    #    deny  all;
    #}
}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220804,220804#msg-220804


From nginx at nginxuser.net  Wed Jan  4 20:01:56 2012
From: nginx at nginxuser.net (Nginx User)
Date: Wed, 4 Jan 2012 23:01:56 +0300
Subject: Is nginx vulnerable to the Hash Table Vulnerability (n.runs AG)?
In-Reply-To: <CAB4Tn6Nb976BK7kvoBW57bysgM4QN3BHGysFny5bh_CZRvBGmg@mail.gmail.com>
References: <CAEHYtVjfNoA6uYQE+DmOFEzj05znaMoNmUGRa6FtOTsD2EOJQQ@mail.gmail.com>
	<CAB4Tn6OU5h3m5GRkZUTynoixJ9MedyXvduhXh4kCNQvkzrc3TQ@mail.gmail.com>
	<1285211291852136747@unknownmsgid>
	<CAB4Tn6PT9ZdEwt3eKzfXYS=wyT-A120RH7R_QgDrs9=a9DVi_A@mail.gmail.com>
	<CAB4Tn6Nb976BK7kvoBW57bysgM4QN3BHGysFny5bh_CZRvBGmg@mail.gmail.com>
Message-ID: <CAN=WzV1U11A=dHtCckp+2cnb03BAUgQsy7LWKA0oYTc13G9=Rg@mail.gmail.com>

On 4 January 2012 14:48, agentzh <agentzh at gmail.com> wrote:
> I've also added similar protections to ngx.req.get_headers():
>
> ? ?http://wiki.nginx.org/HttpLuaModule#ngx.req.get_headers
>
> All of these changes have been released as ngx_lua 0.3.1rc45:
Nice one.

Thanks.


From nginx-forum at nginx.us  Wed Jan  4 20:12:10 2012
From: nginx-forum at nginx.us (piotr)
Date: Wed, 04 Jan 2012 15:12:10 -0500
Subject: proxy_ignore_headers doesn't seem to work
In-Reply-To: <20111218111735.GN67687@mdounin.ru>
References: <20111218111735.GN67687@mdounin.ru>
Message-ID: <205695db0bbc533cbb1587461c79dbf3.NginxMailingListEnglish@forum.nginx.org>

Maxim,

Sorry for the late response.

I'm trying to hide those headers, you can see it in config posted
before:

proxy_ignore_headers X-Accel-Expires Expires Cache-Control Set-Cookie;
proxy_cache_valid 200 302 600h;
proxy_cache_valid any 90m;

But as you can see, you can find expire, cache-control headers in proxy
repsonse, and also in headers of stored cache files on disk. With values
from upstream server, not those set in config.


I'm not sure now those times from my previous response, but whatever i
do (delete cached files from disk, restart nginx, wait 10 minutes to
expire files cached on disk), nginx is always returning in response
cache-control and expire in 600 sec - taking this value from upstream
server, not from the config file. Also, file is stored on disk for 600s,
after this time i can see in logs that nginx is downloading again file
from upstream server. Nginx should cache files for 600h in current
config, but i've also tried before other values, like 6h or 6d etc.

http://mpq1.XXX.com/tiles/1.0.0/osm/12/1035/1519.jpg

GET /tiles/1.0.0/osm/12/1035/1519.jpg HTTP/1.1
Host: mpq1.XXX.com
User-Agent: Mozilla/5.0 (Ubuntu; X11; Linux i686; rv:9.0.1)
Gecko/20100101 Firefox/9.0.1
Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8
Accept-Language: en,en-us;q=0.7,pl;q=0.3
Accept-Encoding: gzip, deflate
Accept-Charset: ISO-8859-2,utf-8;q=0.7,*;q=0.7
Connection: keep-alive
Cookie: bblastvisit=1325618979; bblastactivity=0;
__utma=26892847.843027784.1325618982.1325618982.1325703951.2;
__utmz=26892847.1325618982.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none);
__qca=P0-367095175-1325618982208;
bbsessionhash=02acddda9354e88dd04bab92358aeca6; __utmc=26892847

HTTP/1.1 200 OK
Server: nginx/1.0.11
Date: Wed, 04 Jan 2012 19:52:01 GMT
Content-Type: image/jpeg
Connection: keep-alive
Content-Length: 8010
Cache-Control: max-age=600
Last-Modified: Wed, 04 Jan 2012 02:18:15 GMT
Expires: Wed, 04 Jan 2012 20:02:01 GMT



Regards,
Piotr

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220058,220808#msg-220808


From unai at leanservers.com  Wed Jan  4 21:05:33 2012
From: unai at leanservers.com (Unai Rodriguez)
Date: Thu, 5 Jan 2012 05:05:33 +0800
Subject: Simply change the listen port ,does not work
In-Reply-To: <a88e1717bd60a8d0c468ddf8dccb680c.NginxMailingListEnglish@forum.nginx.org>
References: <a88e1717bd60a8d0c468ddf8dccb680c.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <97645564-55A9-42AD-BBC9-64FB47AA1420@leanservers.com>

Fxofxo,

What do you mean by "nothing happened, nginx failed"? Is there any specific error on the logs? Are you sure NGINX is being restarted properly? Is NGINX listening on port 82 after you restart it?

Unai Rodriguez
CTO & Cofounder
LeanWired LLP
Singapore 168977
www.leanservers.com
unai at leanservers.com

On 5 Jan, 2012, at 3:05, "fxofxo" <nginx-forum at nginx.us> wrote:

> Hi, I just installed nginx-1.0.11 , and start it by default , view
> http://xxx.xxx.xxx.xxx:80  ,and nginx works
> Then I change the listen port in default.conf  to 82,  kill nginx ,
> restart , and view http://xxx.xxx.xxx.xxx:82, nothing happened, nginx 
> failed. 
> Could anyone tell me what's wrong with it ? Thanks very much.
> 
> server {
>    listen       82;
>    server_name  localhost;
> 
>    #charset koi8-r;
>    #access_log  /var/log/nginx/log/host.access.log  main;
> 
>    location / {
>        root   /usr/share/nginx/html;
>        index  index.html index.htm;
>    }
> 
>    #error_page  404              /404.html;
> 
>    # redirect server error pages to the static page /50x.html
>    #
>    error_page   500 502 503 504  /50x.html;
>    location = /50x.html {
>        root   /usr/share/nginx/html;
>    }
> 
>    # 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$ {
>    #    root           html;
>    #    fastcgi_pass   127.0.0.1:9000;
>    #    fastcgi_index  index.php;
>    #    fastcgi_param  SCRIPT_FILENAME  /scripts$fastcgi_script_name;
>    #    include        fastcgi_params;
>    #}
> 
>    # deny access to .htaccess files, if Apache's document root
>    # concurs with nginx's one
>    #
>    #location ~ /\.ht {
>    #    deny  all;
>    #}
> }
> 
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220804,220804#msg-220804
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From francis at daoine.org  Wed Jan  4 22:26:05 2012
From: francis at daoine.org (Francis Daly)
Date: Wed, 4 Jan 2012 22:26:05 +0000
Subject: proxy_ignore_headers doesn't seem to work
In-Reply-To: <205695db0bbc533cbb1587461c79dbf3.NginxMailingListEnglish@forum.nginx.org>
References: <20111218111735.GN67687@mdounin.ru>
	<205695db0bbc533cbb1587461c79dbf3.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120104222605.GT27443@craic.sysops.org>

On Wed, Jan 04, 2012 at 03:12:10PM -0500, piotr wrote:

Hi there,

> I'm trying to hide those headers, you can see it in config posted
> before:
> 
> proxy_ignore_headers X-Accel-Expires Expires Cache-Control Set-Cookie;

No, that's proxy_ignore_headers, not proxy_hide_header.

> proxy_cache_valid 200 302 600h;
> proxy_cache_valid any 90m;
> 
> But as you can see, you can find expire, cache-control headers in proxy
> repsonse, and also in headers of stored cache files on disk. With values
> from upstream server, not those set in config.

What comes from upstream is stored on disk. That's expected.

proxy_ignore_headers determines which of upstream headers or
proxy_cache_valid is used by nginx to decide for how long nginx will
cache the response.

Separate from that, you can use "proxy_hide_header" to tell nginx not to
send some headers that came from upstream, to the client.

Separate from that (mostly), you can use "expires" to tell nginx how to
set Expires: and Cache-Control: headers in the response to the client.

"(mostly)" is there because nginx will only send a single Expires header,
so if you use "expires" to set one, then the one from upstream will not
go to the client, even if it isn't in "proxy_hide_header".

> I'm not sure now those times from my previous response, but whatever i
> do (delete cached files from disk, restart nginx, wait 10 minutes to
> expire files cached on disk), nginx is always returning in response
> cache-control and expire in 600 sec - taking this value from upstream
> server, not from the config file.

That is as expected, since you have no other value in the config file
to send in the response, and you have nothing in the config file to tell
nginx not to send what came from upstream.

> Also, file is stored on disk for 600s,
> after this time i can see in logs that nginx is downloading again file
> from upstream server. Nginx should cache files for 600h in current
> config, but i've also tried before other values, like 6h or 6d etc.

That observation, I don't understand.

You can examine the first few bytes of the on-disk cache file to see
when nginx thinks that it will become stale:

od -t dL $cache_file | head -n 1

The second and fourth numbers there are valid-until- and fetched-at-
timestamps, respectively. So if they are only 600 apart, then something
unexpected (to me) is happening.

When I use configuration like

> proxy_cache_valid 200 302 600h;
> proxy_cache_valid any 90m;

in 1.1.11, I see them being the expected 2160000 (seconds, = 600 hours)
apart for a http 200 response.

Good luck,

	f
-- 
Francis Daly        francis at daoine.org


From nginx-forum at nginx.us  Wed Jan  4 22:30:42 2012
From: nginx-forum at nginx.us (freejack)
Date: Wed, 04 Jan 2012 17:30:42 -0500
Subject: Trailing slash being removed from document root
Message-ID: <a50e717121a71198f1f56e55905d1af5.NginxMailingListEnglish@forum.nginx.org>

Hi all.  We're considering switching from lighttpd to nginx in my
company and so far everything looks good except for one thing I can't
seem to figure out.

When getting the document root in PHP from nginx using this:

$_SERVER['DOCUMENT_ROOT']."../".$ZONE."/".$LOCALE...etc......

nginx will always return the path WITHOUT the trailing slash.  I have
the document root configured like this:

root           /home/web/www.site.com/html/; # note the trailing slash

We have way too much code in php that gets the document root without
adding a trailing slash after $_SERVER['DOCUMENT_ROOT'] because lighttpd
and apache before that always had preserved the trailing slash.

How can I configure nginx to keep the slash?

Thanks!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220811,220811#msg-220811


From francis at daoine.org  Wed Jan  4 22:56:58 2012
From: francis at daoine.org (Francis Daly)
Date: Wed, 4 Jan 2012 22:56:58 +0000
Subject: Trailing slash being removed from document root
In-Reply-To: <a50e717121a71198f1f56e55905d1af5.NginxMailingListEnglish@forum.nginx.org>
References: <a50e717121a71198f1f56e55905d1af5.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120104225658.GU27443@craic.sysops.org>

On Wed, Jan 04, 2012 at 05:30:42PM -0500, freejack wrote:

Hi there,

> When getting the document root in PHP from nginx using this:
> 
> $_SERVER['DOCUMENT_ROOT']."../".$ZONE."/".$LOCALE...etc......
> 
> nginx will always return the path WITHOUT the trailing slash.

nginx doesn't do PHP, it does fastcgi. Which means that it sends a series
of key/value params to the fastcgi server, which in turn presents them
to your PHP in the _SERVER array.

Which in turn means that...

> We have way too much code in php that gets the document root without
> adding a trailing slash after $_SERVER['DOCUMENT_ROOT'] because lighttpd
> and apache before that always had preserved the trailing slash.
> 
> How can I configure nginx to keep the slash?

...wherever you currently set "fastcgi_param DOCUMENT_ROOT" (probably from
"include fastcgi.conf"), add the slash there and it all should Just Work.

That is:

you probably have

  fastcgi_param DOCUMENT_ROOT $document_root;

Change it to be

  fastcgi_param DOCUMENT_ROOT $document_root/;

and on nginx -s reload, you should see the difference.

	f
-- 
Francis Daly        francis at daoine.org


From francis at daoine.org  Wed Jan  4 23:13:26 2012
From: francis at daoine.org (Francis Daly)
Date: Wed, 4 Jan 2012 23:13:26 +0000
Subject: proxy_ignore_headers doesn't seem to work
In-Reply-To: <20120104222605.GT27443@craic.sysops.org>
References: <20111218111735.GN67687@mdounin.ru>
	<205695db0bbc533cbb1587461c79dbf3.NginxMailingListEnglish@forum.nginx.org>
	<20120104222605.GT27443@craic.sysops.org>
Message-ID: <20120104231326.GV27443@craic.sysops.org>

On Wed, Jan 04, 2012 at 10:26:05PM +0000, Francis Daly wrote:
> On Wed, Jan 04, 2012 at 03:12:10PM -0500, piotr wrote:

Hi there,

One more thing on the final "600s"...

> > Also, file is stored on disk for 600s,
> > after this time i can see in logs that nginx is downloading again file
> > from upstream server. Nginx should cache files for 600h in current
> > config, but i've also tried before other values, like 6h or 6d etc.
> 
> That observation, I don't understand.

You don't specify "inactive" in your proxy_cache_path directive, so it
defaults to 10 minutes.

If you see one new request per url every 11 minutes, then it is expected
that nginx will re-fetch from upstream each time.

If you see a new request every 9 minutes, then nginx should not need to
re-fetch from upstream.

Does that explain what you see?

	f
-- 
Francis Daly        francis at daoine.org


From nginx-forum at nginx.us  Wed Jan  4 23:16:36 2012
From: nginx-forum at nginx.us (freejack)
Date: Wed, 04 Jan 2012 18:16:36 -0500
Subject: Trailing slash being removed from document root
In-Reply-To: <a50e717121a71198f1f56e55905d1af5.NginxMailingListEnglish@forum.nginx.org>
References: <a50e717121a71198f1f56e55905d1af5.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <0ff38419352beda9dab73b11dd2ffe32.NginxMailingListEnglish@forum.nginx.org>

Thanks that did the trick!  Still have a lot to learn... ;)

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220811,220814#msg-220814


From stef at caunter.ca  Wed Jan  4 23:31:56 2012
From: stef at caunter.ca (Stefan Caunter)
Date: Wed, 4 Jan 2012 18:31:56 -0500
Subject: Optimize nginx for uploading/download huge files..
In-Reply-To: <aa3060b95f2e21e88b0942423f46407b.NginxMailingListEnglish@forum.nginx.org>
References: <aa3060b95f2e21e88b0942423f46407b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CALdpnrzLX-0TvZgPc3dCQAdnO5os+XzG8B64aH-aNNRzxQmxTw@mail.gmail.com>

On Tue, Jan 3, 2012 at 5:35 PM, ronjpark <nginx-forum at nginx.us> wrote:
> Hello everyone,
>
> Firstly, I'm sorry for posting the same question here and there (How to
> forum).
> It is urgent issue to me.
>
> Could you give me a tip for which directives and values I have to set
> for the optimization?
>
> I installed nginx on virtual machine as reverse proxy server.
> It checks the authentication and bypasses the client request to the
> backend servers.
> The most transactions are to upload/download files, from 1M to 1G...
>
> The CPU usage hits 100% when the 1M file download request reaches around
> 25 TPS..
> around 20 TPS for upload request..
>
> I set client_body_buffer_size, proxy_buffer_size, proxy_buffers and so
> on..
> But not so good yet.. :(
>
> client_body_buffer_size is set to 2M and it helps to increase TPS for
> uploading, but not enough.
> proxy_buffer_size = 1M and proxy_buffers = 4 1M help to increase TPS for
> downloading little bit, but not so much..
> Other directives are set by its default.
>
> The below is the result of profiling with google_proftools_module.
> As you can see, the most time is spent on readv() and writev().
>
> Please give me any advice or your experience.
> Thank you for your time.
>
> Total: 22244 samples
> 11928 53.6% 53.6% 11930 53.6% readv
> 8212 36.9% 90.5% 8212 36.9% writev
> 907 4.1% 94.6% 907 4.1% recv
> 359 1.6% 96.2% 359 1.6% __connect_nocancel
> 248 1.1% 97.3% 248 1.1% __close_nocancel
> 35 0.2% 97.5% 46 0.2% _IO_str_pbackfail
> 34 0.2% 97.7% 34 0.2% __write_nocancel
> 31 0.1% 97.8% 31 0.1% socket
> 28 0.1% 97.9% 28 0.1% memset
> and
> Total: 64475 samples
> 34117 52.9% 52.9% 34117 52.9% writev
> 14173 22.0% 74.9% 14181 22.0% readv
> 11981 18.6% 93.5% 11981 18.6% recv
> 1046 1.6% 95.1% 1049 1.6% __connect_nocancel
> 878 1.4% 96.5% 878 1.4% __close_nocancel
> 343 0.5% 97.0% 343 0.5% epoll_wait
> 160 0.2% 97.2% 160 0.2% __write_nocancel
> 89 0.1% 97.4% 89 0.1% __xstat64
> 81 0.1% 97.5% 115 0.2% _IO_str_pbackfail
>
>
> Here is the information of my VM:
> ronjpark at PSEProxy:~/release/REL.0.13$ cat /proc/cpuinfo
> processor : 0
> vendor_id : GenuineIntel
> cpu family : 6
> model : 2
> model name : QEMU Virtual CPU version 0.14.1
> stepping : 3
> cpu MHz : 2393.998
> cache size : 4096 KB
> fpu : yes
> fpu_exception : yes
> cpuid level : 4
> wp : yes
> flags : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36
> clflush mmx fxsr sse sse2 syscall nx lm up rep_good nopl pni cx16 popcnt
> hypervisor lahf_lm
> bogomips : 4787.99
> clflush size : 64
> cache_alignment : 64
> address sizes : 40 bits physical, 48 bits virtual
> power management:
>
> ronjpark at PSEProxy:~/release/REL.0.13$ cat /proc/meminfo
> MemTotal: 4057052 kB
> MemFree: 2651832 kB
> Buffers: 80984 kB
> Cached: 1100692 kB
> SwapCached: 3856 kB
> Active: 325744 kB
> Inactive: 910176 kB
> Active(anon): 48884 kB
> Inactive(anon): 5684 kB
> Active(file): 276860 kB
> Inactive(file): 904492 kB
> Unevictable: 0 kB
> Mlocked: 0 kB
> SwapTotal: 2095100 kB
> SwapFree: 2082688 kB
> Dirty: 4 kB
> Writeback: 0 kB
> AnonPages: 51356 kB
> Mapped: 15740 kB
> Shmem: 324 kB
> Slab: 66252 kB
> SReclaimable: 56796 kB
> SUnreclaim: 9456 kB
> KernelStack: 1160 kB
> PageTables: 4124 kB
> NFS_Unstable: 0 kB
> Bounce: 0 kB
> WritebackTmp: 0 kB
> CommitLimit: 4123624 kB
> Committed_AS: 176112 kB
> VmallocTotal: 34359738367 kB
> VmallocUsed: 28816 kB
> VmallocChunk: 34359707236 kB
> HardwareCorrupted: 0 kB
> HugePages_Total: 0
> HugePages_Free: 0
> HugePages_Rsvd: 0
> HugePages_Surp: 0
> Hugepagesize: 2048 kB
> DirectMap4k: 42996 kB
> DirectMap2M: 4151296 kB
> ronjpark at PSEProxy:~/release/REL.0.13$ uname -a
> Linux PSEProxy 2.6.38-8-server #42-Ubuntu SMP Mon Apr 11 03:49:04 UTC
> 2011 x86_64 x86_64 x86_64 GNU/Linux

The bottleneck is neither CPU nor memory. VM disk performance will be
bad. Any throughput you get on the virtual disk will degrade as each
concurrent upload client triggers reads and writes.  This is one case
where you need a SAN, ZFS filesystem, or an SSD drive; configure with
large block size if files are >1GB. Even server class hard drives will
thrash and slow down everything with enough clients contending for the
disk. Remember, in relative time, if a cpu operation takes a second,
an operation on the disk takes one month, under ideal conditions, and
you are giving nginx less than ideal conditions.

Stefan Caunter
http://scaleengine.com/contact


From nginx-forum at nginx.us  Wed Jan  4 23:51:16 2012
From: nginx-forum at nginx.us (DeneB)
Date: Wed, 04 Jan 2012 18:51:16 -0500
Subject: NGinX Form Input Problem
Message-ID: <7b47560704a625f5d162e375a3a6171a.NginxMailingListEnglish@forum.nginx.org>

I have just recently set-up a new LEMP in Ubuntu 11.10 (64-bit) and
running NGinX ver 1.0.5. the server is running fine but I cannot seem to
get any forms to work. On my old server (windows) the forms worked fine,
but now I get the following error "502 Bad Gateway".

NGinX was installed with apt-get install, it is not a version that I
compiled myself. Also I am running PHP5-Fastcgi.

I have tried setting the forms to work in two different ways, one to
write to a text file and also to write to MySQL database.....but with no
luck.

Any ideas on how to get the forms working again??

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220816,220816#msg-220816


From nginx-forum at nginx.us  Thu Jan  5 00:38:08 2012
From: nginx-forum at nginx.us (ronjpark)
Date: Wed, 04 Jan 2012 19:38:08 -0500
Subject: Optimize nginx for uploading/download huge files..
In-Reply-To: <CALdpnrzLX-0TvZgPc3dCQAdnO5os+XzG8B64aH-aNNRzxQmxTw@mail.gmail.com>
References: <CALdpnrzLX-0TvZgPc3dCQAdnO5os+XzG8B64aH-aNNRzxQmxTw@mail.gmail.com>
Message-ID: <9c6788ce9aa062102b81fa4ddba7d9d3.NginxMailingListEnglish@forum.nginx.org>

stefancaunter Wrote:
-------------------------------------------------------
> On Tue, Jan 3, 2012 at 5:35 PM, ronjpark
> <nginx-forum at nginx.us> wrote:
> 
> The bottleneck is neither CPU nor memory. VM disk
> performance will be
> bad. Any throughput you get on the virtual disk
> will degrade as each
> concurrent upload client triggers reads and
> writes.  This is one case
> where you need a SAN, ZFS filesystem, or an SSD
> drive; configure with
> large block size if files are >1GB. Even server
> class hard drives will
> thrash and slow down everything with enough
> clients contending for the
> disk. Remember, in relative time, if a cpu
> operation takes a second,
> an operation on the disk takes one month, under
> ideal conditions, and
> you are giving nginx less than ideal conditions.
> 
> Stefan Caunter
> http://scaleengine.com/contact
> 

Thank you Stefan.
I also doubted what you said and tested with 1MB file.

client_body_buffer_size is set to 2M and proxy_buffer_size = 2M also.
But it doesn't work.

As my understainding, nginx doesn't make disk I/O for this file
transaction because of enough big buffer size. Right?
Please fix me if I'm wrong.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220778,220818#msg-220818


From mdounin at mdounin.ru  Thu Jan  5 03:18:57 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 5 Jan 2012 07:18:57 +0400
Subject: proxy_ignore_headers doesn't seem to work
In-Reply-To: <205695db0bbc533cbb1587461c79dbf3.NginxMailingListEnglish@forum.nginx.org>
References: <20111218111735.GN67687@mdounin.ru>
	<205695db0bbc533cbb1587461c79dbf3.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120105031856.GM67687@mdounin.ru>

Hello!

On Wed, Jan 04, 2012 at 03:12:10PM -0500, piotr wrote:

> Maxim,
> 
> Sorry for the late response.
> 
> I'm trying to hide those headers, you can see it in config posted
> before:
> 
> proxy_ignore_headers X-Accel-Expires Expires Cache-Control Set-Cookie;
> proxy_cache_valid 200 302 600h;
> proxy_cache_valid any 90m;
> 
> But as you can see, you can find expire, cache-control headers in proxy
> repsonse, and also in headers of stored cache files on disk. With values
> from upstream server, not those set in config.

You don't hide the headers, and don't set headers in config.  The 
proxy_ignore_headers instruct nginx to ignore headers in question 
(i.e. don't set/limit cache validity time as stored in cache 
metadata based on these headers).  The proxy_cache_valid 
directives set cache validity times.  None of these directives 
hide or modify headers in question.

To hide headers, you have to use the proxy_hide_header directive 
(as already suggested in previous message), see here for details:

http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_hide_header

To explicitly set Cache-Control/Expires headers, use the "expires" 
directive, see here for details:

http://wiki.nginx.org/HttpHeadersModule#expires

(no official docs in English yet)

> I'm not sure now those times from my previous response, but whatever i
> do (delete cached files from disk, restart nginx, wait 10 minutes to
> expire files cached on disk), nginx is always returning in response
> cache-control and expire in 600 sec - taking this value from upstream
> server, not from the config file. Also, file is stored on disk for 600s,
> after this time i can see in logs that nginx is downloading again file
> from upstream server. Nginx should cache files for 600h in current
> config, but i've also tried before other values, like 6h or 6d etc.

Do you request the file in question with interval larger than 600 
seconds, without any intermediate requests?  It is expected to be 
removed from cache as inactive one in such test, see desciption of 
the "inactive" parameter of the "proxy_cache_path" directive here:

http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_cache_path

Try requesting a file once per minute to keep it active.

Maxim Dounin


From nginx-forum at nginx.us  Thu Jan  5 03:45:30 2012
From: nginx-forum at nginx.us (DeneB)
Date: Wed, 04 Jan 2012 22:45:30 -0500
Subject: NGinX Form Input Problem
In-Reply-To: <7b47560704a625f5d162e375a3a6171a.NginxMailingListEnglish@forum.nginx.org>
References: <7b47560704a625f5d162e375a3a6171a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <b713d11a5e56e7a7644c35cd68823843.NginxMailingListEnglish@forum.nginx.org>

So the only thing that I have found is this
HttpFormInputModule.......looks like I will have to uninstall the rpm
version and then download the source code and recompile/reconfigure
nginx.  

The last time I did this I had added the http_flv_module but during that
recompiling I could not get the http_xslt_module to compile at all and
did it without_http_xslt_module.

I really didn't want to do this but as far as the mechanics goes for my
site I am almost there......just need to be able to accept data from
forms.

Does this sound like I am on the right path???

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220816,220820#msg-220820


From unai at leanservers.com  Thu Jan  5 03:49:44 2012
From: unai at leanservers.com (Unai Rodriguez)
Date: Thu, 5 Jan 2012 11:49:44 +0800
Subject: NGinX Form Input Problem
In-Reply-To: <7b47560704a625f5d162e375a3a6171a.NginxMailingListEnglish@forum.nginx.org>
References: <7b47560704a625f5d162e375a3a6171a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <E6BB6E1F651B4A80B7EB5D75DF844670@leanservers.com>

Do you get the error "502 bad gateway" with any PHP page? It seems PHP might not be working at all. I would recommend creating a very simple PHP page and testing if it works with the following content for example:

<?php

phpinfo();

?>

Let us know if this works. 

-- 
Unai Rodriguez
CTO & Cofounder
LeanWired LLP
Singapore 168977
www.leanservers.com
unai at leanservers.com




On Thursday, January 5, 2012 at 7:51 AM, DeneB wrote:

> I have just recently set-up a new LEMP in Ubuntu 11.10 (64-bit) and
> running NGinX ver 1.0.5. the server is running fine but I cannot seem to
> get any forms to work. On my old server (windows) the forms worked fine,
> but now I get the following error "502 Bad Gateway".
> 
> NGinX was installed with apt-get install, it is not a version that I
> compiled myself. Also I am running PHP5-Fastcgi.
> 
> I have tried setting the forms to work in two different ways, one to
> write to a text file and also to write to MySQL database.....but with no
> luck.
> 
> Any ideas on how to get the forms working again??
> 
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220816,220816#msg-220816
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org (mailto:nginx at nginx.org)
> http://mailman.nginx.org/mailman/listinfo/nginx
> 
> 


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120105/7b33a227/attachment.html>

From nginx-forum at nginx.us  Thu Jan  5 04:43:06 2012
From: nginx-forum at nginx.us (DeneB)
Date: Wed, 04 Jan 2012 23:43:06 -0500
Subject: NGinX Form Input Problem
In-Reply-To: <E6BB6E1F651B4A80B7EB5D75DF844670@leanservers.com>
References: <E6BB6E1F651B4A80B7EB5D75DF844670@leanservers.com>
Message-ID: <a0c97889a51f1d61dd9f2b20b5c1532c.NginxMailingListEnglish@forum.nginx.org>

Yes I get the error, "502 Bad Gateway", when I request a PHP page.
I checked on my terminal and php5-fpm is running.

I also use a php script in the nginx config to run webstats which is
also running.

What information should I post to help solve this problem?????

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220816,220822#msg-220822


From nginx-forum at nginx.us  Thu Jan  5 05:44:49 2012
From: nginx-forum at nginx.us (DeneB)
Date: Thu, 05 Jan 2012 00:44:49 -0500
Subject: NGinX Form Input Problem
In-Reply-To: <a0c97889a51f1d61dd9f2b20b5c1532c.NginxMailingListEnglish@forum.nginx.org>
References: <E6BB6E1F651B4A80B7EB5D75DF844670@leanservers.com>
	<a0c97889a51f1d61dd9f2b20b5c1532c.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <dd74783ac6fcfb8df8e2ff940c3f0b27.NginxMailingListEnglish@forum.nginx.org>

okay.....I retraced my installation of PHP and it looks like I had never
started a PHP FastCGI daemon (if that makes any sense) now I am able to
see the php info page.  I can also see under modules that MySQL is
running.  I am going to test my forms again right now.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220816,220824#msg-220824


From nginx-forum at nginx.us  Thu Jan  5 06:14:11 2012
From: nginx-forum at nginx.us (DeneB)
Date: Thu, 05 Jan 2012 01:14:11 -0500
Subject: NGinX Form Input Problem
In-Reply-To: <dd74783ac6fcfb8df8e2ff940c3f0b27.NginxMailingListEnglish@forum.nginx.org>
References: <E6BB6E1F651B4A80B7EB5D75DF844670@leanservers.com>
	<a0c97889a51f1d61dd9f2b20b5c1532c.NginxMailingListEnglish@forum.nginx.org>
	<dd74783ac6fcfb8df8e2ff940c3f0b27.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <68ce260980203090c1d9ff3f3990c05b.NginxMailingListEnglish@forum.nginx.org>

Unai thank-you so much for your help and pointing me in the right
direction.  I just did a test and now the forms are working.  I can
write to a text file and when I changed it to see if I could write to a
database I did not get the 502 BAD Gateway Error.  (Instead I got a
message saying it couldn't connect to the database.....but I think this
is my error.....going to set up a new dBase and test again.)

I must say you folks that hang out here are absolutley brilliant!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220816,220825#msg-220825


From reallfqq-nginx at yahoo.fr  Thu Jan  5 06:33:06 2012
From: reallfqq-nginx at yahoo.fr (B.R.)
Date: Thu, 5 Jan 2012 01:33:06 -0500
Subject: Error log file
Message-ID: <CALqce=3VxJPE86qn6k9SZRLJryB6JJtkkPP96BWSA1Z8qLJLbg@mail.gmail.com>

Hello,

Using Nginx 1.0.11, I changed the error log file in my configuration to
/var/log/nginx.error.log.
The new log file gets filled as it is supposed to be, but the old file
/nginx-repo/logs/error.log keeps getting recreated.

I would add that I checked the configuration so there is no conflicting
entry.

Btw, why does nginx doesn't use the 'standard' repos /var/log for the logs
and /var/run for the PID file by default? It adds to the entropy...
---
*B. R.*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120105/843fadfb/attachment.html>

From nginx-forum at nginx.us  Thu Jan  5 09:14:57 2012
From: nginx-forum at nginx.us (DeneB)
Date: Thu, 05 Jan 2012 04:14:57 -0500
Subject: NGinX Form Input Problem
In-Reply-To: <68ce260980203090c1d9ff3f3990c05b.NginxMailingListEnglish@forum.nginx.org>
References: <E6BB6E1F651B4A80B7EB5D75DF844670@leanservers.com>
	<a0c97889a51f1d61dd9f2b20b5c1532c.NginxMailingListEnglish@forum.nginx.org>
	<dd74783ac6fcfb8df8e2ff940c3f0b27.NginxMailingListEnglish@forum.nginx.org>
	<68ce260980203090c1d9ff3f3990c05b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <5845d95e3dbed8e47ad2d2de2e3fca4b.NginxMailingListEnglish@forum.nginx.org>

Forms are writing to the database!  Thanks Unai, and have a great day!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220816,220828#msg-220828


From unai at leanservers.com  Thu Jan  5 09:25:06 2012
From: unai at leanservers.com (Unai Rodriguez)
Date: Thu, 5 Jan 2012 17:25:06 +0800
Subject: NGinX Form Input Problem
In-Reply-To: <5845d95e3dbed8e47ad2d2de2e3fca4b.NginxMailingListEnglish@forum.nginx.org>
References: <E6BB6E1F651B4A80B7EB5D75DF844670@leanservers.com>
	<a0c97889a51f1d61dd9f2b20b5c1532c.NginxMailingListEnglish@forum.nginx.org>
	<dd74783ac6fcfb8df8e2ff940c3f0b27.NginxMailingListEnglish@forum.nginx.org>
	<68ce260980203090c1d9ff3f3990c05b.NginxMailingListEnglish@forum.nginx.org>
	<5845d95e3dbed8e47ad2d2de2e3fca4b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <6133B03B-0D80-4097-A473-47BBFCD3ACA7@leanservers.com>

Sure, you are very welcome, take care 

Unai Rodriguez
CTO & Cofounder
LeanWired LLP
Singapore 168977
www.leanservers.com
unai at leanservers.com

On 5 Jan, 2012, at 17:14, "DeneB" <nginx-forum at nginx.us> wrote:

> Forms are writing to the database!  Thanks Unai, and have a great day!
> 
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220816,220828#msg-220828
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From tomasz.roda at gmail.com  Thu Jan  5 09:48:39 2012
From: tomasz.roda at gmail.com (Tomasz Roda)
Date: Thu, 5 Jan 2012 10:48:39 +0100
Subject: Problem with instant streaming request body to backend
Message-ID: <CAM_zQbTuXP+s0=YMEaCyuPfqvHNwGywb9sc_uGhSEWdbeHopCA@mail.gmail.com>

Hey,

I am developer in new startup and Nginx is important part of whole
system. Generally it is great load balancer, has many interesting
modules and configuration possibilities but I have one serious problem
which I cannot solve.

Problem is that Nginx for each request collects the whole body of
request and only after that sends request next to backend. I know that
I can theoretically use upload module to omit that issue but we have
got own part of code to handle uploads in way that we need (which must
not be supported by module in load balancer). That issue is big
problem for us because in current load balancer implementation data
from client is unnecessarily "cached" on load balancer (load
balancer's hard drive in fact) and then send to backend instead of
instant streaming from client by load balancer to backend. Main
problem is lower performance (requirement of fast storage on load
balancer servers) and additional requests latency, it is really
noticeable especially for big files.

My question is: is any chance for adding feature of instant data
streaming of request body to backend? If chance is real can we expect
feature implementation soon? I think that it is useful functionallity
and can be used by many Nginx users.

Tomek


From nginx-forum at nginx.us  Thu Jan  5 10:01:17 2012
From: nginx-forum at nginx.us (atrus)
Date: Thu, 05 Jan 2012 05:01:17 -0500
Subject: How to see nginx rps (requests per second)
Message-ID: <18e0180c916acba5477b1b9f31f5a89a.NginxMailingListEnglish@forum.nginx.org>

Hi Guys,

I have enable stub module and location /status/ in nginx.conf 

All I see just :

Active connections: 1 
server accepts handled requests
 478077 478077 478271 
Reading: 0 Writing: 1 Waiting: 0 

How can I know which values is the rps of the nginx server ? 

Thanks guys.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220831,220831#msg-220831


From smallfish.xy at gmail.com  Thu Jan  5 10:02:55 2012
From: smallfish.xy at gmail.com (smallfish)
Date: Thu, 5 Jan 2012 18:02:55 +0800
Subject: How to see nginx rps (requests per second)
In-Reply-To: <18e0180c916acba5477b1b9f31f5a89a.NginxMailingListEnglish@forum.nginx.org>
References: <18e0180c916acba5477b1b9f31f5a89a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CACPpB2xVreOiyRdG6bdc4YBSc95jFieOeTXwvduVirCjiX=GvA@mail.gmail.com>

use ApacheBench(ab), http_load, siege or other performance tools
--
blog: http://chenxiaoyu.org



2012/1/5 atrus <nginx-forum at nginx.us>

> Hi Guys,
>
> I have enable stub module and location /status/ in nginx.conf
>
> All I see just :
>
> Active connections: 1
> server accepts handled requests
>  478077 478077 478271
> Reading: 0 Writing: 1 Waiting: 0
>
> How can I know which values is the rps of the nginx server ?
>
> Thanks guys.
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,220831,220831#msg-220831
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120105/4fcad845/attachment.html>

From nginx-forum at nginx.us  Thu Jan  5 10:17:24 2012
From: nginx-forum at nginx.us (layneuks)
Date: Thu, 05 Jan 2012 05:17:24 -0500
Subject: rewrite N|next flag in apache
Message-ID: <c63378ae95ebd64e8d6bae6c7addbba1.NginxMailingListEnglish@forum.nginx.org>

Hi everyone,

I'm writing here because i have rewrite problems on Nginx. First of all
i will explain to you what result i'm expecting.
For example i have this URL : 
 
http://www.mydomain.com/api/redirect/var1/value1/var2/value2/varX/valueX/?varX=valueX&...

after rewrite i would like to get : 

http://www.mydomain.com/?var1=value1&var2=value2&varX=valueX&varX=valueX...


So i get it works on apache with : 

RewriteRule
^(api/redirect/[^\/]+/[^\/]+/[^\/]+)/([^\/]+)/([^\/]+)([^\?]*)
$1$4/?&$2=$3 [QSA,N]
RewriteRule ^api/redirect/[^\/]+/[^\/]+/([^\/]+)
index.php?service=Redirect&hash=$1 [QSA,L]

And now i would like to have the same on nginx, but i think nginx
doesn't handle something like apache's N flag. Then it's my nginx
rewrite rules, it work once because it doesn't re-execute rewrite while
it's possible.

 rewrite ^/(api/redirect/[^\/]+/[^\/]+/[^\/]+)/([^\/]+)/([^\/]+)([^\?]*)
/$1$4/?&$2=$3;
 rewrite ^/api/redirect/[^\/]+/[^\/]+/([^\/]+)
/public/index.php?service=Redirect&func=Redirect&hash=$1 last;

Do someone have a solution for this please ? I will apreciate a lot.
thanks you.

Best regards,

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220833,220833#msg-220833


From nginx-forum at nginx.us  Thu Jan  5 10:41:13 2012
From: nginx-forum at nginx.us (Yogui)
Date: Thu, 05 Jan 2012 05:41:13 -0500
Subject: How to see nginx rps (requests per second)
In-Reply-To: <18e0180c916acba5477b1b9f31f5a89a.NginxMailingListEnglish@forum.nginx.org>
References: <18e0180c916acba5477b1b9f31f5a89a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <069d41a48775bb48b64c0d019ed25575.NginxMailingListEnglish@forum.nginx.org>

Hi,

I beleive the stub module gives the number of requests since nginx was
last reloaded (or started). To get the number of rps, you have to query
the stub module each second and substract the old value from the current
value.

Regards,

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220831,220834#msg-220834


From andrew at nginx.com  Thu Jan  5 11:53:29 2012
From: andrew at nginx.com (Andrew Alexeev)
Date: Thu, 5 Jan 2012 15:53:29 +0400
Subject: Problem with instant streaming request body to backend
In-Reply-To: <CAM_zQbTuXP+s0=YMEaCyuPfqvHNwGywb9sc_uGhSEWdbeHopCA@mail.gmail.com>
References: <CAM_zQbTuXP+s0=YMEaCyuPfqvHNwGywb9sc_uGhSEWdbeHopCA@mail.gmail.com>
Message-ID: <BBBF4804-9D3F-490F-AF71-70AEE5F9F56F@nginx.com>

Tomek,

On Jan 5, 2012, at 1:48 PM, Tomasz Roda wrote:

> Hey,
> 
> I am developer in new startup and Nginx is important part of whole
> system. Generally it is great load balancer, has many interesting
> modules and configuration possibilities but I have one serious problem
> which I cannot solve.
> 
> Problem is that Nginx for each request collects the whole body of
> request and only after that sends request next to backend. I know that
> I can theoretically use upload module to omit that issue but we have
> got own part of code to handle uploads in way that we need (which must
> not be supported by module in load balancer). That issue is big
> problem for us because in current load balancer implementation data
> from client is unnecessarily "cached" on load balancer (load
> balancer's hard drive in fact) and then send to backend instead of
> instant streaming from client by load balancer to backend. Main
> problem is lower performance (requirement of fast storage on load
> balancer servers) and additional requests latency, it is really
> noticeable especially for big files.
> 
> My question is: is any chance for adding feature of instant data
> streaming of request body to backend? If chance is real can we expect
> feature implementation soon? I think that it is useful functionallity
> and can be used by many Nginx users.

By instant data streaming do you mean generic L4/TCP load balancing or still some form of a L7/HTTP load balancing with additional controls for headers and request/response, and with efficient handling of concurrent connections?

What are the other roles of your nginx setup aside from load balancing to the backends?

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


From tomasz.roda at gmail.com  Thu Jan  5 12:27:53 2012
From: tomasz.roda at gmail.com (Tomasz Roda)
Date: Thu, 5 Jan 2012 13:27:53 +0100
Subject: Problem with instant streaming request body to backend
In-Reply-To: <BBBF4804-9D3F-490F-AF71-70AEE5F9F56F@nginx.com>
References: <CAM_zQbTuXP+s0=YMEaCyuPfqvHNwGywb9sc_uGhSEWdbeHopCA@mail.gmail.com>
	<BBBF4804-9D3F-490F-AF71-70AEE5F9F56F@nginx.com>
Message-ID: <CAM_zQbS7fk5jx960APhYZj-k_8Kpm9sWP9Vygb7sMGOqzz2A-g@mail.gmail.com>

2012/1/5 Andrew Alexeev <andrew at nginx.com>:
> Tomek,
>
> On Jan 5, 2012, at 1:48 PM, Tomasz Roda wrote:
>
>> Hey,
>>
>> I am developer in new startup and Nginx is important part of whole
>> system. Generally it is great load balancer, has many interesting
>> modules and configuration possibilities but I have one serious problem
>> which I cannot solve.
>>
>> Problem is that Nginx for each request collects the whole body of
>> request and only after that sends request next to backend. I know that
>> I can theoretically use upload module to omit that issue but we have
>> got own part of code to handle uploads in way that we need (which must
>> not be supported by module in load balancer). That issue is big
>> problem for us because in current load balancer implementation data
>> from client is unnecessarily "cached" on load balancer (load
>> balancer's hard drive in fact) and then send to backend instead of
>> instant streaming from client by load balancer to backend. Main
>> problem is lower performance (requirement of fast storage on load
>> balancer servers) and additional requests latency, it is really
>> noticeable especially for big files.
>>
>> My question is: is any chance for adding feature of instant data
>> streaming of request body to backend? If chance is real can we expect
>> feature implementation soon? I think that it is useful functionallity
>> and can be used by many Nginx users.
>
> By instant data streaming do you mean generic L4/TCP load balancing or still some form of a L7/HTTP load balancing with additional controls for headers and request/response, and with efficient handling of concurrent connections?
>
> What are the other roles of your nginx setup aside from load balancing to the backends?
>

I mean a L7/HTTP load balancing. We need additional handling of
requests by Nginx for HTTPS (backend applications operate on HTTP
requests), access logging, passing requests to proper backend
applications basing on URL, x-accel-redirect, etc.


From andrew at nginx.com  Thu Jan  5 12:44:41 2012
From: andrew at nginx.com (Andrew Alexeev)
Date: Thu, 5 Jan 2012 16:44:41 +0400
Subject: Problem with instant streaming request body to backend
In-Reply-To: <CAM_zQbS7fk5jx960APhYZj-k_8Kpm9sWP9Vygb7sMGOqzz2A-g@mail.gmail.com>
References: <CAM_zQbTuXP+s0=YMEaCyuPfqvHNwGywb9sc_uGhSEWdbeHopCA@mail.gmail.com>
	<BBBF4804-9D3F-490F-AF71-70AEE5F9F56F@nginx.com>
	<CAM_zQbS7fk5jx960APhYZj-k_8Kpm9sWP9Vygb7sMGOqzz2A-g@mail.gmail.com>
Message-ID: <FC6689F3-DF8F-4BBB-AF69-A3FAE0C3120A@nginx.com>

Tomek,

On Jan 5, 2012, at 4:27 PM, Tomasz Roda wrote:

> 2012/1/5 Andrew Alexeev <andrew at nginx.com>:
>> Tomek,
>> 
>> On Jan 5, 2012, at 1:48 PM, Tomasz Roda wrote:
>> 
>>> Hey,
>>> 
>>> I am developer in new startup and Nginx is important part of whole
>>> system. Generally it is great load balancer, has many interesting
>>> modules and configuration possibilities but I have one serious problem
>>> which I cannot solve.
>>> 
>>> Problem is that Nginx for each request collects the whole body of
>>> request and only after that sends request next to backend. I know that
>>> I can theoretically use upload module to omit that issue but we have
>>> got own part of code to handle uploads in way that we need (which must
>>> not be supported by module in load balancer). That issue is big
>>> problem for us because in current load balancer implementation data
>>> from client is unnecessarily "cached" on load balancer (load
>>> balancer's hard drive in fact) and then send to backend instead of
>>> instant streaming from client by load balancer to backend. Main
>>> problem is lower performance (requirement of fast storage on load
>>> balancer servers) and additional requests latency, it is really
>>> noticeable especially for big files.
>>> 
>>> My question is: is any chance for adding feature of instant data
>>> streaming of request body to backend? If chance is real can we expect
>>> feature implementation soon? I think that it is useful functionallity
>>> and can be used by many Nginx users.
>> 
>> By instant data streaming do you mean generic L4/TCP load balancing or still some form of a L7/HTTP load balancing with additional controls for headers and request/response, and with efficient handling of concurrent connections?
>> 
>> What are the other roles of your nginx setup aside from load balancing to the backends?
>> 
> 
> I mean a L7/HTTP load balancing. We need additional handling of
> requests by Nginx for HTTPS (backend applications operate on HTTP
> requests), access logging, passing requests to proper backend
> applications basing on URL, x-accel-redirect, etc.

There's some work planned towards this direction (allowing to avoid buffering from the client), mostly to facilitate websockets stuff. But probably will turn out to be a generic "unbuffered L4/L7" load balancing. There's no ETA, though and it's probably a three months timeframe.


From mdounin at mdounin.ru  Thu Jan  5 13:04:38 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 5 Jan 2012 17:04:38 +0400
Subject: Optimize nginx for uploading/download huge files..
In-Reply-To: <9c6788ce9aa062102b81fa4ddba7d9d3.NginxMailingListEnglish@forum.nginx.org>
References: <CALdpnrzLX-0TvZgPc3dCQAdnO5os+XzG8B64aH-aNNRzxQmxTw@mail.gmail.com>
	<9c6788ce9aa062102b81fa4ddba7d9d3.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120105130438.GN67687@mdounin.ru>

Hello!

On Wed, Jan 04, 2012 at 07:38:08PM -0500, ronjpark wrote:

> stefancaunter Wrote:
> -------------------------------------------------------
> > On Tue, Jan 3, 2012 at 5:35 PM, ronjpark
> > <nginx-forum at nginx.us> wrote:
> > 
> > The bottleneck is neither CPU nor memory. VM disk
> > performance will be
> > bad. Any throughput you get on the virtual disk
> > will degrade as each
> > concurrent upload client triggers reads and
> > writes.  This is one case
> > where you need a SAN, ZFS filesystem, or an SSD
> > drive; configure with
> > large block size if files are >1GB. Even server
> > class hard drives will
> > thrash and slow down everything with enough
> > clients contending for the
> > disk. Remember, in relative time, if a cpu
> > operation takes a second,
> > an operation on the disk takes one month, under
> > ideal conditions, and
> > you are giving nginx less than ideal conditions.
> > 
> > Stefan Caunter
> > http://scaleengine.com/contact
> > 
> 
> Thank you Stefan.
> I also doubted what you said and tested with 1MB file.
> 
> client_body_buffer_size is set to 2M and proxy_buffer_size = 2M also.
> But it doesn't work.
> 
> As my understainding, nginx doesn't make disk I/O for this file
> transaction because of enough big buffer size. Right?
> Please fix me if I'm wrong.

Yes.

To limit disk I/O for big requests/responses there are following 
options available:

1. Using larger buffers, notably client_body_buffer_size, 
proxy_buffer_size, proxy_buffers.  This is basically what you've 
already done.

2. For responses you may also completely disable disk buffering 
using "proxy_max_temp_file_size 0".  This implies that nginx won't 
be able to read full response from a backend though, and 
connection to a backend will be busy for a time needed to client to 
download a response.

Directive description may be found here:
http://wiki.nginx.org/HttpProxyModule#proxy_max_temp_file_size

Maxim Dounin


From jason at monsterjam.org  Thu Jan  5 13:25:43 2012
From: jason at monsterjam.org (jason at monsterjam.org)
Date: Thu, 5 Jan 2012 08:25:43 -0500
Subject: need help getting perl cgis to work.
Message-ID: <20120105132543.GA17409@monsterjam.org>

hey folks, ive been running apache for years but now trying to use nginx. I have nginx and php(with fastcgi) 
working fine, but whenever I try to bring up my web page that has a perl cgi on it, it just downloads the script 
instead of executing it.. I cant figure out how to make it work right.. basically heres the config that I have 
for it in my apache conf

ScriptAliasMatch ^/blog/?(.*) /var/www/localhost/htdocs/blog/blog.pl/$1

and 

<Directory "/var/www/localhost/htdocs/blog">
        Options +Includes
        <FilesMatch "blog.pl$">
        Options ExecCGI 
        </FilesMatch>

when do I need to do to get nginx to properly execute the perl cgi?

thanks/regards,
Jason


From francis at daoine.org  Thu Jan  5 13:43:12 2012
From: francis at daoine.org (Francis Daly)
Date: Thu, 5 Jan 2012 13:43:12 +0000
Subject: need help getting perl cgis to work.
In-Reply-To: <20120105132543.GA17409@monsterjam.org>
References: <20120105132543.GA17409@monsterjam.org>
Message-ID: <20120105134312.GW27443@craic.sysops.org>

On Thu, Jan 05, 2012 at 08:25:43AM -0500, jason at monsterjam.org wrote:

Hi there,

> hey folks, ive been running apache for years but now trying to use nginx. I have nginx and php(with fastcgi) 
> working fine, but whenever I try to bring up my web page that has a perl cgi on it, it just downloads the script 

> when do I need to do to get nginx to properly execute the perl cgi?

nginx doesn't do cgi.

You'll need to run some separate server that nginx can talk to as a
client, that is able to do whatever processing you want.

If you have a http server that handles the cgi for you, you can
"proxy_pass" to that.

If you have a fastcgi server that is set up to do the right thing with
your cgi script, you can "fastcgi_pass" to that.

There are also other options.

On the nginx side, you configure it in a similar way to your
php-via-fastcgi configuration.

	f
-- 
Francis Daly        francis at daoine.org


From nginx-forum at nginx.us  Thu Jan  5 14:12:26 2012
From: nginx-forum at nginx.us (atrus)
Date: Thu, 05 Jan 2012 09:12:26 -0500
Subject: How to see nginx rps (requests per second)
In-Reply-To: <069d41a48775bb48b64c0d019ed25575.NginxMailingListEnglish@forum.nginx.org>
References: <18e0180c916acba5477b1b9f31f5a89a.NginxMailingListEnglish@forum.nginx.org>
	<069d41a48775bb48b64c0d019ed25575.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <84406ea2a7cd2f84d7adc2e5be3856c1.NginxMailingListEnglish@forum.nginx.org>

Thanks Yogui.

Is there any other ways to do this automatically instead of manually
subtract ?! I want to see in live.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220831,220843#msg-220843


From al-nginx at none.at  Thu Jan  5 14:34:50 2012
From: al-nginx at none.at (Aleksandar Lazic)
Date: Thu, 05 Jan 2012 15:34:50 +0100
Subject: need help getting perl cgis to work.
In-Reply-To: <20120105132543.GA17409@monsterjam.org>
References: <20120105132543.GA17409@monsterjam.org>
Message-ID: <83cdba0aef3dfc8b85f8970716eaa2f2@none.at>

Hi,

I use the

http://localdomain.pl/wiki/FcgiWrap

for mailgraph.cgi behind nginx.

BR
Aleks

On 05-01-2012 14:25, jason at monsterjam.org wrote:
> hey folks, ive been running apache for years but now trying to use
> nginx. I have nginx and php(with fastcgi)
> working fine, but whenever I try to bring up my web page that has a
> perl cgi on it, it just downloads the script
> instead of executing it.. I cant figure out how to make it work
> right.. basically heres the config that I have
> for it in my apache conf
>
> ScriptAliasMatch ^/blog/?(.*) 
> /var/www/localhost/htdocs/blog/blog.pl/$1
>
> and
>
> <Directory "/var/www/localhost/htdocs/blog">
>         Options +Includes
>         <FilesMatch "blog.pl$">
>         Options ExecCGI
>         </FilesMatch>
>
> when do I need to do to get nginx to properly execute the perl cgi?
>
> thanks/regards,
> Jason
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From nginx-forum at nginx.us  Thu Jan  5 15:22:24 2012
From: nginx-forum at nginx.us (layneuks)
Date: Thu, 05 Jan 2012 10:22:24 -0500
Subject: rewrite N|next flag in apache
In-Reply-To: <c63378ae95ebd64e8d6bae6c7addbba1.NginxMailingListEnglish@forum.nginx.org>
References: <c63378ae95ebd64e8d6bae6c7addbba1.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <61a1c23ef92d580621d89752f96728f1.NginxMailingListEnglish@forum.nginx.org>

No one ?  I need a solution... ;'(

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220833,220845#msg-220845


From rcole at ilstech.com  Thu Jan  5 16:36:00 2012
From: rcole at ilstech.com (Ryan Cole)
Date: Thu, 5 Jan 2012 10:36:00 -0600
Subject: Changing worker user account causes throttling of response size?
Message-ID: <91ddb2b0c4c28603c7a31dd17d1a7ae3@mail.gmail.com>

Hi all,



I'm setting up Nginx on a RHEL machine. I am proxying to an
ASP.NETapplication. The
ASP.NET application has some nifty built-in document viewer tool, and so it
often downloads relatively large files (in the realm of 1MB-5MB, or
larger). Nginx, when using the default 'nginx' worker user account, has
been working fine. I recently tried enabling auth_basic, and because of
that tried changing the worker user account to my own personal user
account, so that it can access the htpasswd file from my home directory.
This was working fine as well, until we started getting phone calls about
the document viewer from the ASP.NET application saying it did not receive
the entire document download. I investigated, and it looks like Nginx is
throttling that ASP.NET response -> Nginx proxy -> Web user, after I
changed the worker process user account. To test this, I removed the
auth_basic lines that I added, and left the user account as my personal
account, still. It continued to throttle the response size. It seems to cut
the response off at around 48k pretty consistently - I cannot tell if it's
a time-out or a throttle, but it's generally around 48k of a response
before being cut off. To further test, I switched back to the 'nginx'
worker user account, but added back in the auth_basic lines and the
responses return to normal with no throttling.



So to my untrained eye, it looks like switching this user account is
causing it to pick up some different configuration settings, or something,
and chooses to throttle response sizes. I scanned through all of the
available config files and did not see anything that stood out to me. I
scanned the documentation also, but none of the timeout or throttling
options looked relative to this.



Does anyone have any ideas?



Thanks in advance,

Ryan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120105/9418860b/attachment-0001.html>

From ilan at time4learning.com  Thu Jan  5 17:05:36 2012
From: ilan at time4learning.com (Ilan Berkner)
Date: Thu, 5 Jan 2012 12:05:36 -0500
Subject: 404 not showing up in error logs (for PHP files)
Message-ID: <CAOoWpPpZg8hK_Z9+rOe9vYKUoeV1gaQAnMe8D8H2_1p8sJ6soQ@mail.gmail.com>

Hi All,

When a user enters an invalid php file name they are correctly routed to
our default 404.html:

try_files $fastcgi_script_name =404;

The problem is that the 404 error is not showing up in our error logs.

Other 404 errors do show up in our system.

Please advise.

Thanks!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120105/cc6cd485/attachment.html>

From cliff at develix.com  Thu Jan  5 21:33:52 2012
From: cliff at develix.com (Cliff Wells)
Date: Thu, 05 Jan 2012 13:33:52 -0800
Subject: need help getting perl cgis to work.
In-Reply-To: <20120105132543.GA17409@monsterjam.org>
References: <20120105132543.GA17409@monsterjam.org>
Message-ID: <1325799232.25712.1.camel@portable-evil>

As others have mentioned, Nginx doesn't support CGI.

Please see

 http://wiki.nginx.org/Configuration#CGI 

for a couple different options.

Cliff

On Thu, 2012-01-05 at 08:25 -0500, jason at monsterjam.org wrote:
> hey folks, ive been running apache for years but now trying to use nginx. I have nginx and php(with fastcgi) 
> working fine, but whenever I try to bring up my web page that has a perl cgi on it, it just downloads the script 
> instead of executing it.. I cant figure out how to make it work right.. basically heres the config that I have 
> for it in my apache conf
> 
> ScriptAliasMatch ^/blog/?(.*) /var/www/localhost/htdocs/blog/blog.pl/$1
> 
> and 
> 
> <Directory "/var/www/localhost/htdocs/blog">
>         Options +Includes
>         <FilesMatch "blog.pl$">
>         Options ExecCGI 
>         </FilesMatch>
> 
> when do I need to do to get nginx to properly execute the perl cgi?
> 
> thanks/regards,
> Jason
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx



From ft at falkotimme.com  Thu Jan  5 21:55:22 2012
From: ft at falkotimme.com (Falko Timme)
Date: Thu, 5 Jan 2012 22:55:22 +0100
Subject: need help getting perl cgis to work.
References: <20120105132543.GA17409@monsterjam.org>
Message-ID: <E712BBB4E02341E59638DE240FC71057@notebook>

You can find three different solutions here:

http://www.howtoforge.com/serving-cgi-scripts-with-nginx-on-debian-squeeze-ubuntu-11.04

(I prefer Fcgiwrap.)

----- Original Message ----- 
From: <jason at monsterjam.org>
To: <nginx at nginx.org>
Sent: Thursday, January 05, 2012 2:25 PM
Subject: need help getting perl cgis to work.


> hey folks, ive been running apache for years but now trying to use nginx. 
> I have nginx and php(with fastcgi)
> working fine, but whenever I try to bring up my web page that has a perl 
> cgi on it, it just downloads the script
> instead of executing it.. I cant figure out how to make it work right.. 
> basically heres the config that I have
> for it in my apache conf
>
> ScriptAliasMatch ^/blog/?(.*) /var/www/localhost/htdocs/blog/blog.pl/$1
>
> and
>
> <Directory "/var/www/localhost/htdocs/blog">
>        Options +Includes
>        <FilesMatch "blog.pl$">
>        Options ExecCGI
>        </FilesMatch>
>
> when do I need to do to get nginx to properly execute the perl cgi?
>
> thanks/regards,
> Jason
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx 


From ft at falkotimme.com  Thu Jan  5 22:03:17 2012
From: ft at falkotimme.com (Falko Timme)
Date: Thu, 5 Jan 2012 23:03:17 +0100
Subject: 404 not showing up in error logs (for PHP files)
References: <CAOoWpPpZg8hK_Z9+rOe9vYKUoeV1gaQAnMe8D8H2_1p8sJ6soQ@mail.gmail.com>
Message-ID: <816AF7CE5E5A42259DBCEBB87C9673F5@notebook>

Have you tried

fastcgi_intercept_errors on;

?

----- Original Message ----- 
From: Ilan Berkner
To: nginx at nginx.org
Sent: Thursday, January 05, 2012 6:05 PM
Subject: 404 not showing up in error logs (for PHP files)


Hi All,


When a user enters an invalid php file name they are correctly routed to our 
default 404.html:


try_files $fastcgi_script_name =404;


The problem is that the 404 error is not showing up in our error logs.


Other 404 errors do show up in our system.


Please advise.


Thanks!



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


From francis at daoine.org  Thu Jan  5 23:32:49 2012
From: francis at daoine.org (Francis Daly)
Date: Thu, 5 Jan 2012 23:32:49 +0000
Subject: 404 not showing up in error logs (for PHP files)
In-Reply-To: <CAOoWpPpZg8hK_Z9+rOe9vYKUoeV1gaQAnMe8D8H2_1p8sJ6soQ@mail.gmail.com>
References: <CAOoWpPpZg8hK_Z9+rOe9vYKUoeV1gaQAnMe8D8H2_1p8sJ6soQ@mail.gmail.com>
Message-ID: <20120105233249.GY27443@craic.sysops.org>

On Thu, Jan 05, 2012 at 12:05:36PM -0500, Ilan Berkner wrote:

Hi there,

> When a user enters an invalid php file name they are correctly routed to
> our default 404.html:
> 
> try_files $fastcgi_script_name =404;
> 
> The problem is that the 404 error is not showing up in our error logs.
> 
> Other 404 errors do show up in our system.

I don't see any 404s in error.log, and I do see all 404s in access.log.

Is that different from what you see, or different from what you expect?

(I do see in error.log where the nginx static file handler logs when it
fails to open an expected file; but that's distinct from 404 logging.)

	f
-- 
Francis Daly        francis at daoine.org


From ilan at time4learning.com  Thu Jan  5 23:37:15 2012
From: ilan at time4learning.com (Ilan Berkner)
Date: Thu, 5 Jan 2012 18:37:15 -0500
Subject: 404 not showing up in error logs (for PHP files)
In-Reply-To: <20120105233249.GY27443@craic.sysops.org>
References: <CAOoWpPpZg8hK_Z9+rOe9vYKUoeV1gaQAnMe8D8H2_1p8sJ6soQ@mail.gmail.com>
	<20120105233249.GY27443@craic.sysops.org>
Message-ID: <CAOoWpPqgFQbZYrEkVnr+O=5Q1FVXJFO0d8E9V=Pce61AWdRaOg@mail.gmail.com>

I haven't checked the access logs yet (we have them turned off for now).
 I'm just working through various errors, but I understand what you're
saying.  I'll turn the access logs on and see if I see it in there (I'm
sure I will).

Also thanks to Falko pointing me to the error intercept option, although it
may not be necessary at this time.

Ilan

On Thu, Jan 5, 2012 at 6:32 PM, Francis Daly <francis at daoine.org> wrote:

> On Thu, Jan 05, 2012 at 12:05:36PM -0500, Ilan Berkner wrote:
>
> Hi there,
>
> > When a user enters an invalid php file name they are correctly routed to
> > our default 404.html:
> >
> > try_files $fastcgi_script_name =404;
> >
> > The problem is that the 404 error is not showing up in our error logs.
> >
> > Other 404 errors do show up in our system.
>
> I don't see any 404s in error.log, and I do see all 404s in access.log.
>
> Is that different from what you see, or different from what you expect?
>
> (I do see in error.log where the nginx static file handler logs when it
> fails to open an expected file; but that's distinct from 404 logging.)
>
>        f
> --
> Francis Daly        francis at daoine.org
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>



-- 

Ilan Berkner
Chief Technology Officer
Time4Learning.com

6300 NE 1st Ave., Suite 203
Ft. Lauderdale, FL 33334
(954) 771-0914

<http://www.twitter.com/time4learning> <http://www.facebook.com/Time4Learning>


Time4Learning.com - Online interactive curriculum for home use, PreK-8th
Grade.
Time4Writing.com - Online writing tutorials for high, middle, and
elementary school students.
Time4Learning.net - A forum to chat with parents online about kids,
education, parenting and more.
spellingcity.com - Online vocabulary and spelling activities for teachers,
parents and students.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120105/993967ac/attachment.html>

From edho at myconan.net  Fri Jan  6 02:41:21 2012
From: edho at myconan.net (Edho Arief)
Date: Fri, 6 Jan 2012 09:41:21 +0700
Subject: need help getting perl cgis to work.
In-Reply-To: <E712BBB4E02341E59638DE240FC71057@notebook>
References: <20120105132543.GA17409@monsterjam.org>
	<E712BBB4E02341E59638DE240FC71057@notebook>
Message-ID: <CABTjkKmoFq6i5ZcYmedK1i7Os7UMqp7cf-8vNC=HfvUDVL3J8A@mail.gmail.com>

On Fri, Jan 6, 2012 at 4:55 AM, Falko Timme <ft at falkotimme.com> wrote:
> You can find three different solutions here:
>
> http://www.howtoforge.com/serving-cgi-scripts-with-nginx-on-debian-squeeze-ubuntu-11.04
>

And another one here:

http://projects.unbit.it/uwsgi/wiki/CGI

-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From nginx-forum at nginx.us  Fri Jan  6 07:37:02 2012
From: nginx-forum at nginx.us (gigabyte)
Date: Fri, 06 Jan 2012 02:37:02 -0500
Subject: =?UTF-8?Q?Re=3A_why_my_php_file_in_subdirectory_doesn=27t_work=EF=BC=9F?=
In-Reply-To: <61b7f4ff35ccac81d4708939acb29750.NginxMailingListEnglish@forum.nginx.org>
References: <b7f93170ac1488440681bdfe7111f351.NginxMailingListEnglish@forum.nginx.org>
	<61b7f4ff35ccac81d4708939acb29750.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <f79d77c261871f9b655b22dfc5f042c5.NginxMailingListEnglish@forum.nginx.org>

I have the same problem with subdirectories. My config Is:
set  $sroot /usr/local/www/
 location ~ \.php$ {
       include /usr/local/etc/nginx/fastcgi_params;
       fastcgi_index index.php;
       fastcgi_param SCRIPT_FILENAME $sroot/$fastcgi_script_name ;
       fastcgi_param QUERY_STRING    $query_string;
       fastcgi_pass  127.0.0.1:9000;
       fastcgi_intercept_errors on;
       break;
    }

    location / {
       root $sroot ;
       index index.php;
    }

By the way I think it all happend when I've upgraded from nginx 0.9 to
1.0.11. I think that my original configs had been replaced with new
ones.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,215921,220863#msg-220863


From anil at saog.net  Fri Jan  6 08:52:37 2012
From: anil at saog.net (=?UTF-8?B?QW7EsWwgw4dldGlu?=)
Date: Fri, 06 Jan 2012 10:52:37 +0200
Subject: fastcgi cache bypass empty body
Message-ID: <4F06B655.7000707@saog.net>

Hi All,
currently we are using nginx with fastcgi_cache with php backend. For
some conditions php returns an empty body (that we can not find the
cause) but sends some headers and nginx thinks it is valid. So it stores
the file which results to an empty response for upcoming requests.

Cached response from backend is attached, is there a configuration
option to prevent this?

p.s. I changed the site name to <site> in attachment.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: df3c2d9b53cd7241702f43ecfcdc0596
Type: application/octet-stream
Size: 136 bytes
Desc: not available
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120106/81a67a35/attachment-0001.obj>

From ian at ianhobson.co.uk  Fri Jan  6 10:31:07 2012
From: ian at ianhobson.co.uk (Ian Hobson)
Date: Fri, 06 Jan 2012 10:31:07 +0000
Subject: =?UTF-8?Q?Re=3A_why_my_php_file_in_subdirectory_doesn=27t_work=EF=BC=9F?=
In-Reply-To: <f79d77c261871f9b655b22dfc5f042c5.NginxMailingListEnglish@forum.nginx.org>
References: <b7f93170ac1488440681bdfe7111f351.NginxMailingListEnglish@forum.nginx.org>
	<61b7f4ff35ccac81d4708939acb29750.NginxMailingListEnglish@forum.nginx.org>
	<f79d77c261871f9b655b22dfc5f042c5.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4F06CD6B.9060901@ianhobson.co.uk>

On 06/01/2012 07:37, gigabyte wrote:
>         fastcgi_param SCRIPT_FILENAME $sroot/$fastcgi_script_name ;
I suspect this should be

   fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;


Regards

Ian

-- 
Ian Hobson
31 Sheerwater, Northampton NN3 5HU,
Tel: 01604 513875
Preparing eBooks for Kindle and ePub to give the best reader experience.


From edho at myconan.net  Fri Jan  6 10:36:25 2012
From: edho at myconan.net (Edho Arief)
Date: Fri, 6 Jan 2012 17:36:25 +0700
Subject: =?UTF-8?Q?Re=3A_why_my_php_file_in_subdirectory_doesn=27t_work=EF=BC=9F?=
In-Reply-To: <61b7f4ff35ccac81d4708939acb29750.NginxMailingListEnglish@forum.nginx.org>
References: <b7f93170ac1488440681bdfe7111f351.NginxMailingListEnglish@forum.nginx.org>
	<61b7f4ff35ccac81d4708939acb29750.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CABTjkKn3xz3UuGCt+uR_GN3B=K6iPvGLVj-HTFycaD70-jPDew@mail.gmail.com>

On Mon, Oct 17, 2011 at 11:19 PM, suttles <nginx-forum at nginx.us> wrote:
> Update on my problem.... if the URL is typed this way
> http://www.domain.com/subdirectory it will not pull up the site, but
> when the URL is typed http://www.domain.com/subdirectory/ the site comes
> up immediately. How to do force nginx to put the trailing slash onto the
> subdomains?
>

either

location = /subdirectory {
  rewrite ^ $uri/?$args permanent;
}

or

try_files $uri $uri/ =404;

or

symlink in root instead of redefining root in location /subdirectory/

or

location /subdirectory {
  ...
}
##note that this one kind of not what you expect.
/subdirectorysomething will match this location.

-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From appa at perusio.net  Fri Jan  6 11:20:38 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Fri, 06 Jan 2012 11:20:38 +0000
Subject: fastcgi cache bypass empty body
In-Reply-To: <4F06B655.7000707@saog.net>
References: <4F06B655.7000707@saog.net>
Message-ID: <878vll6prt.wl%appa@perusio.net>

On 6 Jan 2012 08h52 WET, anil at saog.net wrote:

> Hi All, currently we are using nginx with fastcgi_cache with php
> backend. For some conditions php returns an empty body (that we can
> not find the cause) but sends some headers and nginx thinks it is
> valid. So it stores the file which results to an empty response for
> upcoming requests.

Perhaps using the embedded Lua module you can capture the response
and, for example, set 'X-Accel-Expires: 0', header for that request or
set a variable that is evaluated by fastcgi_cache_bypass.

then define:

fastcgi_cache_bypass $no_cache;

Check: http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua 

--- appa


From nginx-forum at nginx.us  Fri Jan  6 11:39:17 2012
From: nginx-forum at nginx.us (gigabyte)
Date: Fri, 06 Jan 2012 06:39:17 -0500
Subject: =?UTF-8?Q?Re=3A_why_my_php_file_in_subdirectory_doesn=27t_work=EF=BC=9F?=
In-Reply-To: <4F06CD6B.9060901@ianhobson.co.uk>
References: <4F06CD6B.9060901@ianhobson.co.uk>
Message-ID: <cef574b4ea8135beb88358692e66ab91.NginxMailingListEnglish@forum.nginx.org>

I've already tried meny variations including this one. In fact I'm using
it right now with some modifications:

    location / {
        root $sroot;
        index  index.php index.html index.htm;
        if (-e $request_filename) {
            break;
        }
        rewrite ^/(.+)$ /index.php?q=$1 last;
    }
     if ($host !~* ^www\.) {
        rewrite ^(.*)$ http://www.$host$1 permanent;
    }

    location = /administrator {
        rewrite ^ $uri/?$args permanent;
    }

    location ~ \.php$ {
        root $sroot;
        fastcgi_pass   127.0.0.1:9000;
        fastcgi_index  index.php;
        fastcgi_param  SCRIPT_FILENAME 
$document_root$fastcgi_script_name;
        include        /usr/local/etc/nginx/fastcgi_params;
    }

    location ~ /\.ht {
        deny  all;
    }

But it keeps redirecting me to document_root instead of
document_root/subdirectory

Also I know that It worked before some one put his hands on my config
files.

hobson42 Wrote:
-------------------------------------------------------
> On 06/01/2012 07:37, gigabyte wrote:
> >         fastcgi_param SCRIPT_FILENAME
> $sroot/$fastcgi_script_name ;
> I suspect this should be
> 
>    fastcgi_param SCRIPT_FILENAME
> $document_root$fastcgi_script_name;
> 
> 
> Regards
> 
> Ian
> 
> -- 
> Ian Hobson
> 31 Sheerwater, Northampton NN3 5HU,
> Tel: 01604 513875
> Preparing eBooks for Kindle and ePub to give the
> best reader experience.
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,215921,220872#msg-220872


From edho at myconan.net  Fri Jan  6 11:40:50 2012
From: edho at myconan.net (Edho Arief)
Date: Fri, 6 Jan 2012 18:40:50 +0700
Subject: =?UTF-8?Q?Re=3A_why_my_php_file_in_subdirectory_doesn=27t_work=EF=BC=9F?=
In-Reply-To: <cef574b4ea8135beb88358692e66ab91.NginxMailingListEnglish@forum.nginx.org>
References: <4F06CD6B.9060901@ianhobson.co.uk>
	<cef574b4ea8135beb88358692e66ab91.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CABTjkKmg4VT5-vOfe8jgD85HV5rNn90Km81jMwxxQXc3Qb4hUQ@mail.gmail.com>

On Fri, Jan 6, 2012 at 6:39 PM, gigabyte <nginx-forum at nginx.us> wrote:
> I've already tried meny variations including this one. In fact I'm using
> it right now with some modifications:
>

Post the full config.

-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From nginx-forum at nginx.us  Fri Jan  6 12:26:08 2012
From: nginx-forum at nginx.us (gigabyte)
Date: Fri, 06 Jan 2012 07:26:08 -0500
Subject: =?UTF-8?Q?Re=3A_why_my_php_file_in_subdirectory_doesn=27t_work=EF=BC=9F?=
In-Reply-To: <CABTjkKmg4VT5-vOfe8jgD85HV5rNn90Km81jMwxxQXc3Qb4hUQ@mail.gmail.com>
References: <CABTjkKmg4VT5-vOfe8jgD85HV5rNn90Km81jMwxxQXc3Qb4hUQ@mail.gmail.com>
Message-ID: <a2484c69181e24cbeb333ca9c46123a9.NginxMailingListEnglish@forum.nginx.org>

Here is my full config
server {

    set $sroot /services/www/site;

    listen      192.168.170.1:80;
    server_name  site.net www.site.net;
    access_log  /var/log/nginx/site.net.access.log  main;
    error_log  /var/log/nginx/site.net.error.log info;

    location = /administrator {
        root $sroot/administrator;
    }

    location / {
        root $sroot;
        index  index.php index.html index.htm;
        if (-e $request_filename) {
            break;
        }
        rewrite ^/(.+)$ /index.php?q=$1 last;
    }
     if ($host !~* ^www\.) {
        rewrite ^(.*)$ http://www.$host$1 permanent;
    }


    location ~ \.php$ {
        root $sroot;
        fastcgi_pass   127.0.0.1:9000;
        fastcgi_index  index.php;
        fastcgi_param  SCRIPT_FILENAME 
$document_root$fastcgi_script_name;
        include        /usr/local/etc/nginx/fastcgi_params;
    }

    location ~ /\.ht {
        deny  all;
    }
}

Edho Arief Wrote:
-------------------------------------------------------
> On Fri, Jan 6, 2012 at 6:39 PM, gigabyte
> <nginx-forum at nginx.us> wrote:
> > I've already tried meny variations including
> this one. In fact I'm using
> > it right now with some modifications:
> >
> 
> Post the full config.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,215921,220874#msg-220874


From karthikr2006 at gmail.com  Fri Jan  6 12:36:39 2012
From: karthikr2006 at gmail.com (karthik raman)
Date: Fri, 6 Jan 2012 12:36:39 +0000 (UTC)
Subject: Invitation to connect on LinkedIn
Message-ID: <544185644.5523079.1325853399983.JavaMail.app@ela4-bed77.prod>

LinkedIn
------------



I'd like to add you to my professional network on LinkedIn.

- karthik

karthik raman
--
India

Confirm that you know karthik raman:
https://www.linkedin.com/e/vmkjfu-gx36vqui-17/isd/5422193304/6kyZ0_i4/?hs=false&tok=2EUlKRgvWkbB41

--
You are receiving Invitation to Connect emails. Click to unsubscribe:
http://www.linkedin.com/e/vmkjfu-gx36vqui-17/qjYwmHPVxmbiVvrSCcWdH6Fp6b/goo/nginx%40nginx%2Eorg/20061/I1886930048_1/?hs=false&tok=3B6PP4QLGkbB41

(c) 2011 LinkedIn Corporation. 2029 Stierlin Ct, Mountain View, CA 94043, USA.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120106/3e4a27ca/attachment.html>

From edho at myconan.net  Fri Jan  6 12:48:13 2012
From: edho at myconan.net (Edho Arief)
Date: Fri, 6 Jan 2012 19:48:13 +0700
Subject: =?UTF-8?Q?Re=3A_why_my_php_file_in_subdirectory_doesn=27t_work=EF=BC=9F?=
In-Reply-To: <a2484c69181e24cbeb333ca9c46123a9.NginxMailingListEnglish@forum.nginx.org>
References: <CABTjkKmg4VT5-vOfe8jgD85HV5rNn90Km81jMwxxQXc3Qb4hUQ@mail.gmail.com>
	<a2484c69181e24cbeb333ca9c46123a9.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CABTjkKkYhAsWoa0QJO+EXXVFFqGXCTO+JyApQJhkN4gX+=x6xQ@mail.gmail.com>

On Fri, Jan 6, 2012 at 7:26 PM, gigabyte <nginx-forum at nginx.us> wrote:
> Here is my full config
> server {
>

Try this instead:

http://pastie.org/3137133


-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From pasik at iki.fi  Fri Jan  6 12:55:05 2012
From: pasik at iki.fi (Pasi =?iso-8859-1?Q?K=E4rkk=E4inen?=)
Date: Fri, 6 Jan 2012 14:55:05 +0200
Subject: Plans to integrate HeadersMore module?
In-Reply-To: <20111004082855.GF12984@reaktio.net>
References: <20111004082855.GF12984@reaktio.net>
Message-ID: <20120106125505.GX12984@reaktio.net>

On Tue, Oct 04, 2011 at 11:28:55AM +0300, Pasi K?rkk?inen wrote:
> Hello,
> 
> Any plans to integrate the HeadersMore module (http://wiki.nginx.org/HttpHeadersMoreModule)
> to standard nginx? 
> 
> It's a very handy tool.. especially when you need to modify headers for responses like 401.
> Just trying to minimize the amount of custom-rpm-building for me ;)
> 

Any comments? 

-- Pasi


From nginx-forum at nginx.us  Fri Jan  6 13:46:21 2012
From: nginx-forum at nginx.us (gigabyte)
Date: Fri, 06 Jan 2012 08:46:21 -0500
Subject: =?UTF-8?Q?Re=3A_why_my_php_file_in_subdirectory_doesn=27t_work=EF=BC=9F?=
In-Reply-To: <CABTjkKkYhAsWoa0QJO+EXXVFFqGXCTO+JyApQJhkN4gX+=x6xQ@mail.gmail.com>
References: <CABTjkKkYhAsWoa0QJO+EXXVFFqGXCTO+JyApQJhkN4gX+=x6xQ@mail.gmail.com>
Message-ID: <505cd9fcebc4dec1d5d22943f8c2d7ea.NginxMailingListEnglish@forum.nginx.org>

No, this did not work either.
nginx keeps looking for for relative files in a <root> directory, but
has to do it in <root>/administrator directory
Maybe some other suggestions ?
Edho Arief Wrote:
-------------------------------------------------------
> On Fri, Jan 6, 2012 at 7:26 PM, gigabyte
> <nginx-forum at nginx.us> wrote:
> > Here is my full config
> > server {
> >
> 
> Try this instead:
> 
> http://pastie.org/3137133
> 
> 
> -- 
> O< ascii ribbon campaign - stop html mail -
> www.asciiribbon.org
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,215921,220882#msg-220882


From edho at myconan.net  Fri Jan  6 13:48:25 2012
From: edho at myconan.net (Edho Arief)
Date: Fri, 6 Jan 2012 20:48:25 +0700
Subject: =?UTF-8?Q?Re=3A_why_my_php_file_in_subdirectory_doesn=27t_work=EF=BC=9F?=
In-Reply-To: <505cd9fcebc4dec1d5d22943f8c2d7ea.NginxMailingListEnglish@forum.nginx.org>
References: <CABTjkKkYhAsWoa0QJO+EXXVFFqGXCTO+JyApQJhkN4gX+=x6xQ@mail.gmail.com>
	<505cd9fcebc4dec1d5d22943f8c2d7ea.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CABTjkK=FDE_zqfG9CWbUnLYBDKUk7VeL20LcrJ5OnM+1mHSAxQ@mail.gmail.com>

On Fri, Jan 6, 2012 at 8:46 PM, gigabyte <nginx-forum at nginx.us> wrote:
> No, this did not work either.
> nginx keeps looking for for relative files in a <root> directory, but
> has to do it in <root>/administrator directory
> Maybe some other suggestions ?

What do you mean? Are the files located in
/services/www/site/administrator/administrator/?


-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From brian at akins.org  Fri Jan  6 13:53:35 2012
From: brian at akins.org (Brian Akins)
Date: Fri, 6 Jan 2012 08:53:35 -0500
Subject: Plans to integrate HeadersMore module?
In-Reply-To: <20120106125505.GX12984@reaktio.net>
References: <20111004082855.GF12984@reaktio.net>
	<20120106125505.GX12984@reaktio.net>
Message-ID: <6446A66E-EB06-4802-AD18-70BB7A90C884@akins.org>


On Jan 6, 2012, at 7:55 AM, Pasi K?rkk?inen wrote:

>> 
>> Just trying to minimize the amount of custom-rpm-building for me ;)
>> 
> 
> Any comments? 

We build our own packages using openresty as a base.  Even if openresty, for example, was all in core, we'd still build our own packages. I'm sure others would as well.  For us, it's a simple shell script that does the builds for us, so not a ton of work anyway.  

Digression:

That's one reason I'm not worried about nginx's lack of "dynamic" modules - we'd have to build the modules anyway which would be about as much work as rebuilding nginx, so not really a huge win for us.  We do a good bit of our nginx work in Lua nowadays (thanks agentzh and chaoslawful!)

--Brian


From nginx-forum at nginx.us  Fri Jan  6 14:01:56 2012
From: nginx-forum at nginx.us (gigabyte)
Date: Fri, 06 Jan 2012 09:01:56 -0500
Subject: =?UTF-8?Q?Re=3A_why_my_php_file_in_subdirectory_doesn=27t_work=EF=BC=9F?=
In-Reply-To: <CABTjkK=FDE_zqfG9CWbUnLYBDKUk7VeL20LcrJ5OnM+1mHSAxQ@mail.gmail.com>
References: <CABTjkK=FDE_zqfG9CWbUnLYBDKUk7VeL20LcrJ5OnM+1mHSAxQ@mail.gmail.com>
Message-ID: <28ff7a7da4b72684f09f6df398e076ae.NginxMailingListEnglish@forum.nginx.org>

the Root directory is /services/www/site/

When I login into the index.php which is located in
/services/www/site/administrator
nginx redirects me to index.php which is located in the root directory -
 /services/www/site

When I manually enter site.net/administrator/index.php it opens the
right file but all relative links on it (css,js, ,...etc) are being
loaded from root directory (/services/www/site) and thay should be
loaded from subdirectory (/services/www/site/administrator) and that is
why I'm getting many 404 responses from server.


Edho Arief Wrote:
-------------------------------------------------------
> On Fri, Jan 6, 2012 at 8:46 PM, gigabyte
> <nginx-forum at nginx.us> wrote:
> > No, this did not work either.
> > nginx keeps looking for for relative files in a
> <root> directory, but
> > has to do it in <root>/administrator directory
> > Maybe some other suggestions ?
> 
> What do you mean? Are the files located in
> /services/www/site/administrator/administrator/?
> 
> 
> -- 
> O< ascii ribbon campaign - stop html mail -
> www.asciiribbon.org
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,215921,220885#msg-220885


From brian at akins.org  Fri Jan  6 14:03:58 2012
From: brian at akins.org (Brian Akins)
Date: Fri, 6 Jan 2012 09:03:58 -0500
Subject: How to see nginx rps (requests per second)
In-Reply-To: <84406ea2a7cd2f84d7adc2e5be3856c1.NginxMailingListEnglish@forum.nginx.org>
References: <18e0180c916acba5477b1b9f31f5a89a.NginxMailingListEnglish@forum.nginx.org>
	<069d41a48775bb48b64c0d019ed25575.NginxMailingListEnglish@forum.nginx.org>
	<84406ea2a7cd2f84d7adc2e5be3856c1.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <C497C4EE-B2A0-4F76-9738-7130D2A93BBA@akins.org>


On Jan 5, 2012, at 9:12 AM, atrus wrote:
> 
> Is there any other ways to do this automatically instead of manually
> subtract ?! I want to see in live.

we have some command line ruby/perl scripts that do this, pretty simple.  If you wanted to do this "in" nginx, I guess the easiest way would be using Lua that did a location_capture of stub_status and cached the value ( using http://wiki.nginx.org/HttpLuaModule#ngx.shared.DICT maybe).

Or, have a script that polls nginx stub status every x seconds and writes it out to a static file that nginx can serve.

I'd rather nginx just provide the raw data and I can slice and dice it for my own needs.

--Brian


From edho at myconan.net  Fri Jan  6 14:07:21 2012
From: edho at myconan.net (Edho Arief)
Date: Fri, 6 Jan 2012 21:07:21 +0700
Subject: =?UTF-8?Q?Re=3A_why_my_php_file_in_subdirectory_doesn=27t_work=EF=BC=9F?=
In-Reply-To: <28ff7a7da4b72684f09f6df398e076ae.NginxMailingListEnglish@forum.nginx.org>
References: <CABTjkK=FDE_zqfG9CWbUnLYBDKUk7VeL20LcrJ5OnM+1mHSAxQ@mail.gmail.com>
	<28ff7a7da4b72684f09f6df398e076ae.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CABTjkKkKLyc4QCAHyv=xd3-9Om_r9Pcmg8UV+0LG2TCQSs_3_g@mail.gmail.com>

On Fri, Jan 6, 2012 at 9:01 PM, gigabyte <nginx-forum at nginx.us> wrote:
> the Root directory is /services/www/site/
>
> When I login into the index.php which is located in
> /services/www/site/administrator
> nginx redirects me to index.php which is located in the root directory -
> ?/services/www/site
>
> When I manually enter site.net/administrator/index.php it opens the
> right file but all relative links on it (css,js, ,...etc) are being
> loaded from root directory (/services/www/site) and thay should be
> loaded from subdirectory (/services/www/site/administrator) and that is
> why I'm getting many 404 responses from server.
>
>

Looks like application configuration problem.

-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From nginx-forum at nginx.us  Fri Jan  6 14:23:52 2012
From: nginx-forum at nginx.us (gigabyte)
Date: Fri, 06 Jan 2012 09:23:52 -0500
Subject: =?UTF-8?Q?Re=3A_why_my_php_file_in_subdirectory_doesn=27t_work=EF=BC=9F?=
In-Reply-To: <CABTjkKkKLyc4QCAHyv=xd3-9Om_r9Pcmg8UV+0LG2TCQSs_3_g@mail.gmail.com>
References: <CABTjkKkKLyc4QCAHyv=xd3-9Om_r9Pcmg8UV+0LG2TCQSs_3_g@mail.gmail.com>
Message-ID: <ad8a3c4dc47440b0949ba95f344a790d.NginxMailingListEnglish@forum.nginx.org>

May be but this all started when I've updated nginx from 1.0.4 to 1.0.9
This update erased all my previous config files and I've started from
beginning 
and stuck on this problem
Edho Arief Wrote:
-------------------------------------------------------
> On Fri, Jan 6, 2012 at 9:01 PM, gigabyte
> <nginx-forum at nginx.us> wrote:
> > the Root directory is /services/www/site/
> >
> > When I login into the index.php which is located
> in
> > /services/www/site/administrator
> > nginx redirects me to index.php which is located
> in the root directory -
> > ?/services/www/site
> >
> > When I manually enter
> site.net/administrator/index.php it opens the
> > right file but all relative links on it (css,js,
> ,...etc) are being
> > loaded from root directory (/services/www/site)
> and thay should be
> > loaded from subdirectory
> (/services/www/site/administrator) and that is
> > why I'm getting many 404 responses from server.
> >
> >
> 
> Looks like application configuration problem.
> 
> -- 
> O< ascii ribbon campaign - stop html mail -
> www.asciiribbon.org
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,215921,220888#msg-220888


From cabbar at gmail.com  Fri Jan  6 15:26:59 2012
From: cabbar at gmail.com (Cabbar Duzayak)
Date: Fri, 6 Jan 2012 17:26:59 +0200
Subject: Empty /error log lines (a lot of them)
In-Reply-To: <CAFWPQt6vuXEp7hWg_t=qqEAN0yEcGFa-o0QjJz82Dj0womAOJA@mail.gmail.com>
References: <CA+w7OpK_uNxMz4-1dfCL25HwmmjDe0HyfmV3OBVfvOGZFDrP5g@mail.gmail.com>
	<4F03B1AD.4090207@team.mail.de>
	<CA+w7OpL2C1cbPV-KpQhTuJHPMTkYDC_+AOD+_7cVSjs=CH_1_g@mail.gmail.com>
	<CAFWPQt6vuXEp7hWg_t=qqEAN0yEcGFa-o0QjJz82Dj0womAOJA@mail.gmail.com>
Message-ID: <CA+w7Op+fGu=ZCOpSpakkYJ9=D07PSqUjNNN5V5Zh1Smisv032g@mail.gmail.com>

This worked great!

Thanks a lot for your help.

On Wed, Jan 4, 2012 at 8:27 PM, Mit Rowe <mit at stagename.com> wrote:
>
>
> On Wed, Jan 4, 2012 at 2:03 AM, Cabbar Duzayak <cabbar at gmail.com> wrote:
>>
>> Hi,
>>
>
> [..snip..]
>
>>
>> Can we at least tell Nginx not to log these? Something like, if there
>> is no http method / url, skip the log?
>>
>
>
> It's very likely that there's no Host header sent, so it's being directed to
> the server set as "default".
>
> In my setup i define a server{} section for every real host, and then add a
> separate "catch all" one for everything else and turn off the logs when not
> debugging.
> If you are able to do something similar, it should capture those spurious
> connections, direct them to the default host, and not log the lines
>
> #catch all
> ? ? server {
> ? ? ? ? listen 80 default_server;
> ? ? ? ? server_name localhost.domain.com localhost 127.0.0.1 xx.xx.xx.xx;
> #where xx.xx.xx.xx is the local IP
> ? ? ? ? access_log off;
> ? ? ? ? location / {
> ? ? ? ? ? ? #local
> ? ? ? ? ? ? allow 127.0.0.1;
> ? ? ? ? ? ? #office router
> ? ? ? ? ? ? allow xx.xx.xx.xx;
> ? ? ? ? ? ? #protect
> ? ? ? ? ? ? deny all;
> ? ? ? ? }
> ? ? }
>
> #real server(s)
> ? ?server {
> ? ? ? access_log /var/log/nginx/access.log;
> ? ? ? listen 80;
> ? ? ? server_name .domain.com; #matches www.domain.com and domain.com
> ? ? ?location / {
> ? ? ? ? #[..snip..]
> ? ? ?}
> ? ?}
>
>
>
>
> --
> Will 'Mit' Rowe
> Stagename
> 1-866-326-3098
> mit at stagename.com
> www.stagename.com
> Twitter: @stagename
>
> The information transmitted is intended only for the person or entity to
> which it is addressed and may contain confidential and/or privileged
> material. Any review, retransmission, dissemination or other use of this
> information by persons or entities other than the intended recipient is
> prohibited. If you received this transmission in error, please contact the
> sender and delete all material contained herein from your computer.
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From nginx-forum at nginx.us  Fri Jan  6 20:36:57 2012
From: nginx-forum at nginx.us (kriswpl)
Date: Fri, 06 Jan 2012 15:36:57 -0500
Subject: How to force SNI only connections,
	or have a fallback non-SNI server?
In-Reply-To: <20100714200525.GA80274@rambler-co.ru>
References: <20100714200525.GA80274@rambler-co.ru>
Message-ID: <c3855f2978bc214d8775dabbaf08a50d.NginxMailingListEnglish@forum.nginx.org>

I have the same problem to solve right now. And I see the good idea to
sort the type of SNI and non-SNI supported browsers.

But I made up that the user can connect to the server thru HTTP (no
SSL), and server determine what kind of browser it is (SNI or non) and
depending on the answer redirects to SSL on 443 (default https) for SNI
and 4433 for non-SNI.

So does anyone see how to determine if the client browser support SNI?

I am not talking about a bunch of ifs (browser.version < ... or
browser.version < ...).
I need more something like TLS+SNI support - yes|no - maybe this kind of
combination could tell us - ssl_protocols		SSLv3 TLSv1;

Thanks,
Chris

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,108417,220893#msg-220893


From nginx-forum at nginx.us  Fri Jan  6 21:49:16 2012
From: nginx-forum at nginx.us (gtuhl)
Date: Fri, 06 Jan 2012 16:49:16 -0500
Subject: Nginx as Load Balancer Connection Issues
Message-ID: <20451453746402827e0cb164c69a780b.NginxMailingListEnglish@forum.nginx.org>

We have a box running nginx and two boxes running apache.  The apache
boxes are configured as an upstream for nginx.  

The nginx box has a public IP, and then it talks to the upstream apaches
using the private network (same switch).  We are sustaining a couple
hundred requests/sec.

We've had several issues with the upstreams being counted out by nginx,
causing the "no live upstreams" message in the error log and end users
seeing 502 errors.  When this happens the machines are barely being
used, single digit load averages in 16 core boxes.

Initially we were seeing a ton of "connect() failed (110: Connection
timed out)", 1 every couple seconds.  I added these to sysctl.conf and
that seemed to solve the problem:

net.ipv4.tcp_syncookies = 1
net.ipv4.tcp_fin_timeout = 20    
net.ipv4.tcp_max_syn_backlog = 20480
net.core.netdev_max_backlog = 4096
net.ipv4.tcp_max_tw_buckets = 400000
net.core.somaxconn = 4096

Now things generally run fine but every once in awhile we get a huge
burst of "upstream prematurely closed connection while reading response
header from upstream" followed by a "no live upstreams".  Again, no
apparent load on the machines involved.  These bursts only last a minute
or so.  We also still get an occasional "connect() failed (110:
Connection timed out)" but they are far less frequent, perhaps 1 or 2
per hour.

Anyone have recommendations for tuning the networking side to improve
the situation here?  These are some of the nginx.conf settings we have
in place, removed the ones that don't seem related to the issue:

worker_processes  4;
worker_rlimit_nofile 30000;
events {
    worker_connections  4096;
    # multi_accept on;                                                  
                   
    use epoll;
}
http {
    client_max_body_size 200m;

    proxy_read_timeout 600s;
    proxy_send_timeout 600s;
    proxy_connect_timeout 60s;

    proxy_buffer_size 128k;
    proxy_buffers 4 128k;

    keepalive_timeout  0;
    tcp_nodelay        on;
}

Happy to provide any other details.  This is the "ulimit -a" on all
boxes:

core file size          (blocks, -c) 0
data seg size           (kbytes, -d) unlimited
scheduling priority             (-e) 20
file size               (blocks, -f) unlimited
pending signals                 (-i) 16382
max locked memory       (kbytes, -l) 64
max memory size         (kbytes, -m) unlimited
open files                      (-n) 300000
pipe size            (512 bytes, -p) 8
POSIX message queues     (bytes, -q) 819200
real-time priority              (-r) 0
stack size              (kbytes, -s) 8192
cpu time               (seconds, -t) unlimited
max user processes              (-u) unlimited
virtual memory          (kbytes, -v) unlimited
file locks                      (-x) unlimited

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220894,220894#msg-220894


From nginx-forum at nginx.us  Sat Jan  7 00:17:42 2012
From: nginx-forum at nginx.us (starmonche)
Date: Fri, 06 Jan 2012 19:17:42 -0500
Subject: reverse proxy root folder only
Message-ID: <5170f7a3fb61c2f6067041772f82057a.NginxMailingListEnglish@forum.nginx.org>

I have an nginx serverset up and I'm looking for a way to only reverse
proxy the contents of the root folder. I'm using a location regex with a
list of the files in root that are allowed and then a second catch-all
location that goes nowhere but there must be an easier or cleaner way to
do this.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220895,220895#msg-220895


From nginx-forum at nginx.us  Sat Jan  7 03:12:20 2012
From: nginx-forum at nginx.us (speedfirst)
Date: Fri, 06 Jan 2012 22:12:20 -0500
Subject: How to load a TRUSTED Client CA
Message-ID: <583354cfbe5f6e29a5c0ce36aab40bee.NginxMailingListEnglish@forum.nginx.org>

I need to specify a CA file to "ssl_client_certificate" directive. This
crt is generated by openssl x509 command with "-trustout" parameter, so
starts with "----BEGIN TRUSTED CERTIFICATE-----", rather than common
"-----BEGIN CERTIFICATE-----".

Nginx will report error:
PEM routines:PEM_read_bio:no start line

Internally, nginx uses openssl's API "SSL_load_client_CA_file()" to load
the cert and report this error.

So it's a bug, expected or I mis-configure somewhere?


Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220897,220897#msg-220897


From francis at daoine.org  Sat Jan  7 10:02:27 2012
From: francis at daoine.org (Francis Daly)
Date: Sat, 7 Jan 2012 10:02:27 +0000
Subject: reverse proxy root folder only
In-Reply-To: <5170f7a3fb61c2f6067041772f82057a.NginxMailingListEnglish@forum.nginx.org>
References: <5170f7a3fb61c2f6067041772f82057a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120107100227.GZ27443@craic.sysops.org>

On Fri, Jan 06, 2012 at 07:17:42PM -0500, starmonche wrote:

Hi there,

> I have an nginx serverset up and I'm looking for a way to only reverse
> proxy the contents of the root folder. I'm using a location regex with a
> list of the files in root that are allowed and then a second catch-all
> location that goes nowhere but there must be an easier or cleaner way to
> do this.

Conceptually, one location{} for the things you want, plus one for all
else, *is* the easy way to do it.

(That's more or less the same as one location{} for the things you don't
want, plus one for all else.)

Arguably the cleaner way is to have one exact location{} for each thing
you want, plus one for the rest, but I guess you're not looking for that.

So: what is "the contents of the root folder"? If it is "exactly these
file names", then your current

 location ~ ^/(one|two|three)$

is correct. If it is "any request without a second /", then

 location ~ ^/[^/]*$

should work.

And if you go the other way, and have "location /" as your "this is
the root folder" block, then "all else" would be "any request with two
slashes", which could be

 location ~ /.*/

(All untested, by the way.)

Good luck with it,

	f
-- 
Francis Daly        francis at daoine.org


From mdounin at mdounin.ru  Sat Jan  7 11:37:09 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sat, 7 Jan 2012 15:37:09 +0400
Subject: 404 not showing up in error logs (for PHP files)
In-Reply-To: <CAOoWpPpZg8hK_Z9+rOe9vYKUoeV1gaQAnMe8D8H2_1p8sJ6soQ@mail.gmail.com>
References: <CAOoWpPpZg8hK_Z9+rOe9vYKUoeV1gaQAnMe8D8H2_1p8sJ6soQ@mail.gmail.com>
Message-ID: <20120107113709.GQ67687@mdounin.ru>

Hello!

On Thu, Jan 05, 2012 at 12:05:36PM -0500, Ilan Berkner wrote:

> Hi All,
> 
> When a user enters an invalid php file name they are correctly routed to
> our default 404.html:
> 
> try_files $fastcgi_script_name =404;
> 
> The problem is that the 404 error is not showing up in our error logs.
> 
> Other 404 errors do show up in our system.
> 
> Please advise.

This is expected behaviour.  The "try_files" directives checks 
files as configured and uses configured fallback if there are no 
files found.  It is not expected to log anything to error log 
if there are no files found.

Maxim Dounin


From nginx-forum at nginx.us  Sat Jan  7 13:49:32 2012
From: nginx-forum at nginx.us (AndriesLouw)
Date: Sat, 07 Jan 2012 08:49:32 -0500
Subject: Mobile browser detection at server level
In-Reply-To: <28646c8afe0455c277b2e3a8a7b0c004.squirrel@www.digitalhit.com>
References: <28646c8afe0455c277b2e3a8a7b0c004.squirrel@www.digitalhit.com>
Message-ID: <dc2b6b914cbd74dc83e18b6b37a4c792.NginxMailingListEnglish@forum.nginx.org>

It is perfectly possible to do so with nginx server level configuration,
I wrote a blogpost about it on my blog:
http://blog.andrieslouw.nl/2012/01/serving-mobile-users-with-nginx.html

I'm not exactly sure if it follows all the best-practices, but it seems
to work quite well in my enviroment.

--
Andries Louw Wolthuizen

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,183827,220904#msg-220904


From nginx-forum at nginx.us  Sat Jan  7 13:55:33 2012
From: nginx-forum at nginx.us (AndriesLouw)
Date: Sat, 07 Jan 2012 08:55:33 -0500
Subject: How to see nginx rps (requests per second)
In-Reply-To: <18e0180c916acba5477b1b9f31f5a89a.NginxMailingListEnglish@forum.nginx.org>
References: <18e0180c916acba5477b1b9f31f5a89a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <a3ca201141d32e7bc14e12b9e0074b60.NginxMailingListEnglish@forum.nginx.org>

If you've logging enabled in some sort, you could use a Visual Log
Analyzer, I know some of them work for Apache-style logs, but it
requires the overhead of having logging enabled for your webserver.

Such log-analysers don't need many resources, as they don't require any
interaction with requests served by nginx, they just read the logs nginx
produces every x seconds.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220831,220905#msg-220905


From nginx-forum at nginx.us  Sat Jan  7 14:00:53 2012
From: nginx-forum at nginx.us (AndriesLouw)
Date: Sat, 07 Jan 2012 09:00:53 -0500
Subject: rewrite rule custom.php to custom.css
In-Reply-To: <448df5c84ba71ba759f9506d759831e1.NginxMailingListEnglish@forum.nginx.org>
References: <448df5c84ba71ba759f9506d759831e1.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <3bfeb5929ad566bd2d44f66ee50a9738.NginxMailingListEnglish@forum.nginx.org>

If you want it mod_rewrite style (so no 302 Temporary Redirect header):

location ^/wp-content/themes/mytheme/assets/css/ {
    rewrite custom.css custom.php;
}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220792,220906#msg-220906


From appa at perusio.net  Sat Jan  7 14:25:18 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Sat, 07 Jan 2012 14:25:18 +0000
Subject: Mobile browser detection at server level
In-Reply-To: <dc2b6b914cbd74dc83e18b6b37a4c792.NginxMailingListEnglish@forum.nginx.org>
References: <28646c8afe0455c277b2e3a8a7b0c004.squirrel@www.digitalhit.com>
	<dc2b6b914cbd74dc83e18b6b37a4c792.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87pqev4mk1.wl%appa@perusio.net>

On 7 Jan 2012 13h49 WET, nginx-forum at nginx.us wrote:

> It is perfectly possible to do so with nginx server level
> configuration, I wrote a blogpost about it on my blog:
> http://blog.andrieslouw.nl/2012/01/serving-mobile-users-with-nginx.html
>
> I'm not exactly sure if it follows all the best-practices, but it
> seems to work quite well in my enviroment.

There's a simpler and more efficient way to do it. Instead of
detecting the mobile browser, do the complementary, i.e. detect the
desktop browser:

https://gist.github.com/1326701

--- appa


From nginx-forum at nginx.us  Sat Jan  7 14:35:55 2012
From: nginx-forum at nginx.us (Yogui)
Date: Sat, 07 Jan 2012 09:35:55 -0500
Subject: How to see nginx rps (requests per second)
In-Reply-To: <a3ca201141d32e7bc14e12b9e0074b60.NginxMailingListEnglish@forum.nginx.org>
References: <18e0180c916acba5477b1b9f31f5a89a.NginxMailingListEnglish@forum.nginx.org>
	<a3ca201141d32e7bc14e12b9e0074b60.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <8642603cb607a551ba905c788d88d84b.NginxMailingListEnglish@forum.nginx.org>

@atrus

Please remember that what you are looking for is a "human" value, as
opposed to the snapshot type of values the server can reliably provide.
The reason why nginx does not provide what you are looking for may be
that the server would have to provide several intervals to satisfy
everyone (second, millisecond, minute, hour, etc.), and that is neither
practical nor reliable. Such data could also be misleading, depending on
your use case. From a server point of view, the best thing to do was to
give a total number of requests served and let the admins do the math as
they see fit.

However, the stub status module provides easy-to-read data that you can
parse with a script in a loop, or such a script could dump the stats in
a database and you could then read that. That is the best long-term
solution to your problem. Please remember too that this database does
not need to be SQL, it could be RRD-based for example, or just plain
files, whatever suits you best.

Regards,

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220831,220908#msg-220908


From nginx-forum at nginx.us  Sat Jan  7 14:56:20 2012
From: nginx-forum at nginx.us (AndriesLouw)
Date: Sat, 07 Jan 2012 09:56:20 -0500
Subject: Mobile browser detection at server level
In-Reply-To: <87pqev4mk1.wl%appa@perusio.net>
References: <87pqev4mk1.wl%appa@perusio.net>
Message-ID: <cdcae381b658a48e3d48239b6e58b353.NginxMailingListEnglish@forum.nginx.org>

Ant?nio P. P. Almeida Wrote:
-------------------------------------------------------

> There's a simpler and more efficient way to do it.
> Instead of
> detecting the mobile browser, do the
> complementary, i.e. detect the
> desktop browser:
> 
> https://gist.github.com/1326701
> 
> --- appa

I'd like to warn you about that one, it detects bot's like Googlebot
also as being an mobile device. My setup does also detect desktop
browsers first, as you may or may not have seen. So it's the same setup,
but mine has more exclusions to the "desktop" list.

--
Andries Louw Wolthuizen

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,183827,220909#msg-220909


From brian at akins.org  Sat Jan  7 15:09:49 2012
From: brian at akins.org (Brian Akins)
Date: Sat, 7 Jan 2012 10:09:49 -0500
Subject: Mobile browser detection at server level
In-Reply-To: <cdcae381b658a48e3d48239b6e58b353.NginxMailingListEnglish@forum.nginx.org>
References: <87pqev4mk1.wl%appa@perusio.net>
	<cdcae381b658a48e3d48239b6e58b353.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <963131E6-0B01-4DCD-9D0C-2D33F84D6821@akins.org>

We use some Lua that's basically a combination of the posted links.  It catches the edge cases a little better.  Some things are just easier to do in a "real" programming language.

--Brian

From sconger at brand.net  Fri Jan  6 23:02:54 2012
From: sconger at brand.net (Scott Conger)
Date: Fri, 6 Jan 2012 15:02:54 -0800
Subject: fastcgi_keep_conn consuming ports?
Message-ID: <918068D0CE3C4A48880F53D3EF19861919630D1875@BRANDNETEXCH.Brandnet.local>

I wanted to try the new "fastcgi_keep_conn on;" configuration setting. I grabbed 1.1.13 and added the setting to my configuration. It seems to function, but I started getting HTTP 502 errors when I put load on the server. Looking at netstat, it seems to be consuming all available ports:

netstat -nt > stat.txt
grep TIME_WAIT stat.txt | wc -l
49894

Everything works fine if I comment the configuration line out, which leaves me a bit puzzled as to what is going on. Is there perhaps some other setting I should be changing?

nginx -V
nginx version: nginx/1.1.13
built by gcc 4.1.2 20080704 (Red Hat 4.1.2-50)

-Scott
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120106/59c45bf0/attachment.html>

From for-gmane at mutluit.com  Sat Jan  7 10:09:54 2012
From: for-gmane at mutluit.com (U.Mutlu)
Date: Sat, 07 Jan 2012 11:09:54 +0100
Subject: Disabling logging of absolute paths
Message-ID: <je95lj$86g$1@dough.gmane.org>

Is there a switch to suppress the logging of absolute paths in the log files?
Only paths relative to the html directory should be logged.
Reason: I don't want disclose my file system structure when posting log excerpts to abuse desks etc.

Sanitized example:
2012/01/02 00:04:48 [error] 566#0: *966 open() "/bla/foo/nginx/html/admin/index.php" failed (2: No such file or directory), client: 111.171.207.252, server: localhost, request: "GET //admin/index.php HTTP/1.1", host: "x.200.43.136"

Ie. the "/bla/foo/nginx/html" part should not be logged.


From brian at akins.org  Sat Jan  7 17:46:13 2012
From: brian at akins.org (Brian Akins)
Date: Sat, 7 Jan 2012 12:46:13 -0500
Subject: Disabling logging of absolute paths
In-Reply-To: <je95lj$86g$1@dough.gmane.org>
References: <je95lj$86g$1@dough.gmane.org>
Message-ID: <2806FC5D-91C3-4941-B920-7E70D8AD61BE@akins.org>


On Jan 7, 2012, at 5:09 AM, U.Mutlu wrote:
> 
> Reason: I don't want disclose my file system structure when posting log excerpts to abuse desks etc.

Then redact them before posting.

--Brian


From lists at ruby-forum.com  Sat Jan  7 18:19:33 2012
From: lists at ruby-forum.com (dhaval d.)
Date: Sat, 07 Jan 2012 19:19:33 +0100
Subject: How to see nginx rps (requests per second)
In-Reply-To: <18e0180c916acba5477b1b9f31f5a89a.NginxMailingListEnglish@forum.nginx.org>
References: <18e0180c916acba5477b1b9f31f5a89a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <ee64198b87ba85d94037a6ce828f894f@ruby-forum.com>

atrus wrote in post #1039525:
> Hi Guys,
>
> I have enable stub module and location /status/ in nginx.conf
>
> All I see just :
>
> Active connections: 1
> server accepts handled requests
>  478077 478077 478271
> Reading: 0 Writing: 1 Waiting: 0
>
> How can I know which values is the rps of the nginx server ?
>
> Thanks guys.
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,220831,220831#msg-220831

If you are just interested in knowing Request per second and if you have 
logging enabled then I would suggest using netcat. you just need to tail 
the log file and then use the netcat to count the lines and it would 
give you the details you need. Since the tail by default updates every 1 
second you will get what you need.

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


From nginx-forum at nginx.us  Sat Jan  7 19:11:08 2012
From: nginx-forum at nginx.us (etrader)
Date: Sat, 07 Jan 2012 14:11:08 -0500
Subject: Rewrite for subdomain
Message-ID: <a83d189bc533b95b15f7b184bc24adbe.NginxMailingListEnglish@forum.nginx.org>

I want to set a rewrite rule for redirection of search.php?q=word to
word.domain.com
I used the following rule but did not work.

rewrite ^(.*).$server_name /search.php?q=$1 permanent;

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220917,220917#msg-220917


From mdounin at mdounin.ru  Sat Jan  7 20:32:08 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sun, 8 Jan 2012 00:32:08 +0400
Subject: fastcgi_keep_conn consuming ports?
In-Reply-To: <918068D0CE3C4A48880F53D3EF19861919630D1875@BRANDNETEXCH.Brandnet.local>
References: <918068D0CE3C4A48880F53D3EF19861919630D1875@BRANDNETEXCH.Brandnet.local>
Message-ID: <20120107203208.GR67687@mdounin.ru>

Hello!

On Fri, Jan 06, 2012 at 03:02:54PM -0800, Scott Conger wrote:

> I wanted to try the new "fastcgi_keep_conn on;" configuration 
> setting. I grabbed 1.1.13 and added the setting to my 
> configuration. It seems to function, but I started getting HTTP 
> 502 errors when I put load on the server. Looking at netstat, it 
> seems to be consuming all available ports:
> 
> netstat -nt > stat.txt
> grep TIME_WAIT stat.txt | wc -l
> 49894
> 
> Everything works fine if I comment the configuration line out, 
> which leaves me a bit puzzled as to what is going on. Is there 
> perhaps some other setting I should be changing?
> 
> nginx -V
> nginx version: nginx/1.1.13
> built by gcc 4.1.2 20080704 (Red Hat 4.1.2-50)

If you want to use persistent fastcgi connections, you should use 
fastcgi_keep_conn *and* keepalive directive in the upstream block, 
i.e.

    upstream backend {
        server ...;
        keepalive 5;
    }

    server {
        ...

        location / {
            fastcgi_pass backend;
            fastcgi_keep_conn on;
            ...
        }
    }

Using fastcgi_keep_conn by itself only moves responsibility of 
closing fastcgi connections from backend to nginx, and thus moves 
TIME_WAIT connections from backend(s) to nginx side as well.

Maxim Dounin


From mdounin at mdounin.ru  Sat Jan  7 20:57:24 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sun, 8 Jan 2012 00:57:24 +0400
Subject: Disabling logging of absolute paths
In-Reply-To: <je95lj$86g$1@dough.gmane.org>
References: <je95lj$86g$1@dough.gmane.org>
Message-ID: <20120107205724.GS67687@mdounin.ru>

Hello!

On Sat, Jan 07, 2012 at 11:09:54AM +0100, U.Mutlu wrote:

> Is there a switch to suppress the logging of absolute paths in the log files?
> Only paths relative to the html directory should be logged.
> Reason: I don't want disclose my file system structure when posting log excerpts to abuse desks etc.
> 
> Sanitized example:
> 2012/01/02 00:04:48 [error] 566#0: *966 open() "/bla/foo/nginx/html/admin/index.php" failed (2: No such file or directory), client: 111.171.207.252, server: localhost, request: "GET //admin/index.php HTTP/1.1", host: "x.200.43.136"
> 
> Ie. the "/bla/foo/nginx/html" part should not be logged.

No, there is no such option for error log.  You may use access log 
instead.

Maxim Dounin


From appa at perusio.net  Sat Jan  7 22:47:09 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Sat, 07 Jan 2012 22:47:09 +0000
Subject: rewrite rule custom.php to custom.css
In-Reply-To: <3bfeb5929ad566bd2d44f66ee50a9738.NginxMailingListEnglish@forum.nginx.org>
References: <448df5c84ba71ba759f9506d759831e1.NginxMailingListEnglish@forum.nginx.org>
	<3bfeb5929ad566bd2d44f66ee50a9738.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87mx9z3zbm.wl%appa@perusio.net>

On 7 Jan 2012 14h00 WET, nginx-forum at nginx.us wrote:

> If you want it mod_rewrite style (so no 302 Temporary Redirect
> header):
>
> location ^/wp-content/themes/mytheme/assets/css/ {
> rewrite custom.css custom.php;
> }

Why this? Nginx is not Apache. There's no need for rewrites, in fact
in most mod_rewrite -> nginx conversions you won't need them at all,
IMO.

If you want to return a 301 then just replace the 302 by 301.

Furthermore the '^/' doesn't function since there's no ~ or ~* to denote
a regex based location.

--- appa


From appa at perusio.net  Sat Jan  7 22:51:11 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Sat, 07 Jan 2012 22:51:11 +0000
Subject: Rewrite for subdomain
In-Reply-To: <a83d189bc533b95b15f7b184bc24adbe.NginxMailingListEnglish@forum.nginx.org>
References: <a83d189bc533b95b15f7b184bc24adbe.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87lipj3z4w.wl%appa@perusio.net>

On 7 Jan 2012 19h11 WET, nginx-forum at nginx.us wrote:

> I want to set a rewrite rule for redirection of search.php?q=word to
> word.domain.com
> I used the following rule but did not work.

Try:

location ^~ /search.php {
   return 301 http://$arg_q.domain.com;       
}

--- appa


From edho at myconan.net  Sat Jan  7 23:00:33 2012
From: edho at myconan.net (Edho Arief)
Date: Sun, 8 Jan 2012 06:00:33 +0700
Subject: Rewrite for subdomain
In-Reply-To: <87lipj3z4w.wl%appa@perusio.net>
References: <a83d189bc533b95b15f7b184bc24adbe.NginxMailingListEnglish@forum.nginx.org>
	<87lipj3z4w.wl%appa@perusio.net>
Message-ID: <CABTjkKk8VKUCEFDAjRz5ShAV5xscHHE8Y9wjKU=9TjurTqXEOQ@mail.gmail.com>

On Sun, Jan 8, 2012 at 5:51 AM, Ant?nio P. P. Almeida <appa at perusio.net> wrote:
> On 7 Jan 2012 19h11 WET, nginx-forum at nginx.us wrote:
>
>> I want to set a rewrite rule for redirection of search.php?q=word to
>> word.domain.com
>> I used the following rule but did not work.
>
> Try:
>
> location ^~ /search.php {
> ? return 301 http://$arg_q.domain.com;
> }
>

and the reverse:

server {
  listen 80;
  server_name ~^(?<keyword>.+)\.domain\.com$;
  rewrite ^ http://domain.com/search.php?q=$keyword permanent;
}


-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From nginx-forum at nginx.us  Sat Jan  7 23:08:09 2012
From: nginx-forum at nginx.us (locojohn)
Date: Sat, 07 Jan 2012 18:08:09 -0500
Subject: fastcgi_keep_conn consuming ports?
In-Reply-To: <918068D0CE3C4A48880F53D3EF19861919630D1875@BRANDNETEXCH.Brandnet.local>
References: <918068D0CE3C4A48880F53D3EF19861919630D1875@BRANDNETEXCH.Brandnet.local>
Message-ID: <ca1192fc0d5e483409b37332d51f61f7.NginxMailingListEnglish@forum.nginx.org>

Hello Maxim,

Is there any place where both directives are well documented?  I tried
wiki and the new http://nginx.org/en/docs/, but no luck so far..

Appreciated!

Andrejs

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220911,220923#msg-220923


From mdounin at mdounin.ru  Sat Jan  7 23:11:22 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sun, 8 Jan 2012 03:11:22 +0400
Subject: How to load a TRUSTED Client CA
In-Reply-To: <583354cfbe5f6e29a5c0ce36aab40bee.NginxMailingListEnglish@forum.nginx.org>
References: <583354cfbe5f6e29a5c0ce36aab40bee.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120107231122.GT67687@mdounin.ru>

Hello!

On Fri, Jan 06, 2012 at 10:12:20PM -0500, speedfirst wrote:

> I need to specify a CA file to "ssl_client_certificate" directive. This
> crt is generated by openssl x509 command with "-trustout" parameter, so
> starts with "----BEGIN TRUSTED CERTIFICATE-----", rather than common
> "-----BEGIN CERTIFICATE-----".
> 
> Nginx will report error:
> PEM routines:PEM_read_bio:no start line
> 
> Internally, nginx uses openssl's API "SSL_load_client_CA_file()" to load
> the cert and report this error.
> 
> So it's a bug, expected or I mis-configure somewhere?

I suspect this is the result of the whole trust thing being 
experimental in OpenSSL.

OpenSSL's s_server sample server won't load such certificates as 
well.  It uses the same SSL_load_client_CA_file() function to 
handle -CAfile argument.  Note though that it won't complain as it 
lacks error checking after the call, you should explicitly test 
whether it was loaded by checking CA names list on connect.  On 
the other hand, "openssl verify" is able to handle such certs, so 
it's probably expected to work.  You may try asking OpenSSL's guys 
for details.

Easiest solution probably is to convert the cert to a normal one, 
something like this will do the trick:

$ openssl x509 -in trusted.cert.pem -clrtrust -out normal.cert.pem

Maxim Dounin


From mdounin at mdounin.ru  Sun Jan  8 00:03:39 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sun, 8 Jan 2012 04:03:39 +0400
Subject: fastcgi_keep_conn consuming ports?
In-Reply-To: <ca1192fc0d5e483409b37332d51f61f7.NginxMailingListEnglish@forum.nginx.org>
References: <918068D0CE3C4A48880F53D3EF19861919630D1875@BRANDNETEXCH.Brandnet.local>
	<ca1192fc0d5e483409b37332d51f61f7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120108000339.GU67687@mdounin.ru>

Hello!

On Sat, Jan 07, 2012 at 06:08:09PM -0500, locojohn wrote:

> Hello Maxim,
> 
> Is there any place where both directives are well documented?  I tried
> wiki and the new http://nginx.org/en/docs/, but no luck so far..

Official documentation is expected to appear soon on nginx.org.  

Maxim Dounin


From nginx-forum at nginx.us  Sun Jan  8 00:34:35 2012
From: nginx-forum at nginx.us (Qzi)
Date: Sat, 07 Jan 2012 19:34:35 -0500
Subject: tomcat servlet error ?
Message-ID: <8c0b3f1d46ee87008c5ed68adec80b8a.NginxMailingListEnglish@forum.nginx.org>

I have deploy the webapps of tomcat in the
/usr/share/nginx/html/tomcat/webapps
but when I try it , a error occur.

http://ip/examples/servlets/

nginx error!
The page you are looking for is not found.

but all the jsp and view code page are ok ! only servlet can't work ?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220926,220926#msg-220926


From adam.landas at chinanetcloud.com  Sun Jan  8 05:20:22 2012
From: adam.landas at chinanetcloud.com (Adam Landas)
Date: Sun, 8 Jan 2012 13:20:22 +0800
Subject: tomcat servlet error ?
In-Reply-To: <8c0b3f1d46ee87008c5ed68adec80b8a.NginxMailingListEnglish@forum.nginx.org>
References: <8c0b3f1d46ee87008c5ed68adec80b8a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CAOFyQ6RtfmcmJskkX=Ff=uenAZKN-fmEm0DmYnC5k5A9xc7sdw@mail.gmail.com>

Could you post your config? Are you using nginx as a reverse proxy?


Regards,

Adam
Adam LANDAS | Operations | ChinaNetCloud | www.ChinaNetCloud.com
Phone: +86 (21) 6422-1946 | adam.landas at chinanetcloud.com | Skype:
adamlandas
X2 Space 1-601, 1238 Xietu Lu, Shanghai 200032, China
OnDemand 2011 "100 Top Private Company" - We are hiring!
www.chinanetcloud.com/jobs



On Sun, Jan 8, 2012 at 8:34 AM, Qzi <nginx-forum at nginx.us> wrote:

> I have deploy the webapps of tomcat in the
> /usr/share/nginx/html/tomcat/webapps
> but when I try it , a error occur.
>
> http://ip/examples/servlets/
>
> nginx error!
> The page you are looking for is not found.
>
> but all the jsp and view code page are ok ! only servlet can't work ?
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,220926,220926#msg-220926
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120108/da9c421d/attachment-0001.html>

From mdounin at mdounin.ru  Sun Jan  8 12:49:17 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sun, 8 Jan 2012 16:49:17 +0400
Subject: fastcgi cache bypass empty body
In-Reply-To: <4F06B655.7000707@saog.net>
References: <4F06B655.7000707@saog.net>
Message-ID: <20120108124917.GX67687@mdounin.ru>

Hello!

On Fri, Jan 06, 2012 at 10:52:37AM +0200, An?l ?etin wrote:

> Hi All,
> currently we are using nginx with fastcgi_cache with php backend. For
> some conditions php returns an empty body (that we can not find the
> cause) but sends some headers and nginx thinks it is valid. So it stores
> the file which results to an empty response for upcoming requests.
> 
> Cached response from backend is attached, is there a configuration
> option to prevent this?

No, there is no such option.

Maxim Dounin


From nginx-forum at nginx.us  Sun Jan  8 18:16:57 2012
From: nginx-forum at nginx.us (DenisTRUFFAUT)
Date: Sun, 08 Jan 2012 13:16:57 -0500
Subject: NginX RewriteMap : How to ?
Message-ID: <36e5aa0fc3c1d5f6d429e21441c5c228.NginxMailingListEnglish@forum.nginx.org>

Hi,

I'm wondering how can I emulate the popular Apache's RewriteMap
feature.

I tried several things, without success :

http://forum.nginx.org/read.php?11,220713,220936#msg-220936

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220939,220939#msg-220939


From martinloy.uy at gmail.com  Sun Jan  8 21:14:44 2012
From: martinloy.uy at gmail.com (Martin Loy)
Date: Sun, 8 Jan 2012 19:14:44 -0200
Subject: NginX RewriteMap : How to ?
In-Reply-To: <36e5aa0fc3c1d5f6d429e21441c5c228.NginxMailingListEnglish@forum.nginx.org>
References: <36e5aa0fc3c1d5f6d429e21441c5c228.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CAHbGeW4PgA2mBDUX0h7gmoYA5g_w8iaxGN3ud2ip7W8s8=QwrQ@mail.gmail.com>

Look at http://wiki.nginx.org/HttpMapModule and
http://wiki.nginx.org/HttpRewriteModule

Regards

M

<http://wiki.nginx.org/HttpMapModule>

On Sun, Jan 8, 2012 at 4:16 PM, DenisTRUFFAUT <nginx-forum at nginx.us> wrote:

> Hi,
>
> I'm wondering how can I emulate the popular Apache's RewriteMap
> feature.
>
> I tried several things, without success :
>
> http://forum.nginx.org/read.php?11,220713,220936#msg-220936
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,220939,220939#msg-220939
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>



-- 
*Nunca hubo un amigo que hiciese un favor a un enano, ni un enemigo que le
hiciese un mal, que no se viese recompensado por entero.*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120108/107bb660/attachment.html>

From nginx-forum at nginx.us  Mon Jan  9 08:00:13 2012
From: nginx-forum at nginx.us (cn_nginxer)
Date: Mon, 09 Jan 2012 03:00:13 -0500
Subject: How to Windows auth working on nginx reverse proxy ???
In-Reply-To: <v2s5d7f07421004091450j2abc96c4s3a05dafbbf59071a@mail.gmail.com>
References: <v2s5d7f07421004091450j2abc96c4s3a05dafbbf59071a@mail.gmail.com>
Message-ID: <58a7e2f7e71dd70b4ea1a70e266e0e66.NginxMailingListEnglish@forum.nginx.org>

Hello Ryan,

May I ask an very stupide question, what do you mean by nginx is not
supported back-end keep-alive?

I checked with wireshark, both NTLM and MD authentication are using the
same TCP connection, as far as I understand, nginx is just do a forward
job, so the question is that why nginx could foward the MD request to
the client but not NTLM request?

I hope my dull mind will not bother you so much :-)

Many thanks in advance!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,72871,220949#msg-220949


From nginx-forum at nginx.us  Mon Jan  9 09:17:27 2012
From: nginx-forum at nginx.us (niraj)
Date: Mon, 09 Jan 2012 04:17:27 -0500
Subject: Error during configuration of nginx with html and cgi-bin
Message-ID: <cc0678e4d5b5db9bec4a1c7a99bdd077.NginxMailingListEnglish@forum.nginx.org>

Hi


   I am configuring the nginx , I have added the following entry in
nginx.conf


server {
        listen       80;
        server_name  localhost;
    include       /etc/nginx/mime.types;
    default_type  application/octet-stream;

 location / {
             root  /var/www/vendx/html_asp;
             index  index.html index.htm;
            }

    location ~ \.cgi$ {
    gzip off;
    include /etc/nginx/fastcgi_params;
    fastcgi_pass  unix:/var/run/fcgiwrap.socket;
    fastcgi_index index.cgi;
    fastcgi_param SCRIPT_FILENAME  /var/www/vendx$fastcgi_script_name;
     }




When I trying to login I am getting the following error in logs.


2012/01/09 14:42:29 [error] 21061#0: *1 open()
"/var/www/vendx/html_asp/graphs/mis/amcharts/ampie/data/1326100348_21063_account_qcs_waiting.xml"
failed (2: No such file or directory), client: 192.168.1.44, server:
localhost, request: "GET
/graphs/mis/amcharts/ampie/data/1326100348_21063_account_qcs_waiting.xml
HTTP/1.1", host: "192.168.1.45"

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220951,220951#msg-220951


From nginx-forum at nginx.us  Mon Jan  9 11:29:44 2012
From: nginx-forum at nginx.us (supportnti)
Date: Mon, 09 Jan 2012 06:29:44 -0500
Subject: what is ngx_cycle_t represent
Message-ID: <4de6b93c927cc1d98238cfb189b9a239.NginxMailingListEnglish@forum.nginx.org>

I am new to nginx and trying to understand its internals. What does
ngx_cycle_t represent and why does it have ? 

Appreciate and pointers.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220954,220954#msg-220954


From nginx-forum at nginx.us  Mon Jan  9 11:30:57 2012
From: nginx-forum at nginx.us (supportnti)
Date: Mon, 09 Jan 2012 06:30:57 -0500
Subject: what is ngx_cycle_t represent
In-Reply-To: <4de6b93c927cc1d98238cfb189b9a239.NginxMailingListEnglish@forum.nginx.org>
References: <4de6b93c927cc1d98238cfb189b9a239.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <15227990255e8c136a67e9e8772f63d2.NginxMailingListEnglish@forum.nginx.org>

I am new to nginx and trying to understand its internals. What does
ngx_cycle_t represent?

Appreciate any pointers.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220954,220955#msg-220955


From al-nginx at none.at  Mon Jan  9 12:09:04 2012
From: al-nginx at none.at (Aleksandar Lazic)
Date: Mon, 09 Jan 2012 13:09:04 +0100
Subject: what is ngx_cycle_t represent
In-Reply-To: <4de6b93c927cc1d98238cfb189b9a239.NginxMailingListEnglish@forum.nginx.org>
References: <4de6b93c927cc1d98238cfb189b9a239.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <d97ba48596c42adca7ac39d360f52e87@none.at>

Hi Supportnti,

On 09-01-2012 12:29, supportnti wrote:
> I am new to nginx and trying to understand its internals. What does
> ngx_cycle_t represent and why does it have ?
>
> Appreciate and pointers.

http://lxr.evanmiller.org/http/ident?i=ngx_cycle_t

BR
Aleks


From mdounin at mdounin.ru  Mon Jan  9 14:06:15 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 9 Jan 2012 18:06:15 +0400
Subject: How to Windows auth working on nginx reverse proxy ???
In-Reply-To: <58a7e2f7e71dd70b4ea1a70e266e0e66.NginxMailingListEnglish@forum.nginx.org>
References: <v2s5d7f07421004091450j2abc96c4s3a05dafbbf59071a@mail.gmail.com>
	<58a7e2f7e71dd70b4ea1a70e266e0e66.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120109140615.GE67687@mdounin.ru>

Hello!

On Mon, Jan 09, 2012 at 03:00:13AM -0500, cn_nginxer wrote:

> Hello Ryan,
> 
> May I ask an very stupide question, what do you mean by nginx is not
> supported back-end keep-alive?

Backend keepalive http connections are supported since nginx 1.1.4, but 
it doesn't help to proxy NTLM.  See below.

> I checked with wireshark, both NTLM and MD authentication are using the
> same TCP connection, as far as I understand, nginx is just do a forward
> job, so the question is that why nginx could foward the MD request to
> the client but not NTLM request?

NTLM is connection-based and requires the same connection to be 
used for other requests from the client.  It's not going to work 
through http proxy servers as "the same connection for the same 
client" isn't guaranteed by http proxy servers.  See here for 
details:

http://www.microsoft.com/technet/prodtechnol/WindowsServer2003/Library/IIS/523ae943-5e6a-4200-9103-9808baa00157.mspx?mfr=true

Digest authentication, in contrast, doesn't rely on the same 
connection to be used.  Instead, client provides data originally 
supplied by the server (nonce value) in each request.  See here 
for details:

http://en.wikipedia.org/wiki/Digest_access_authentication
http://tools.ietf.org/html/rfc2617

Maxim Dounin


From nginx-forum at nginx.us  Mon Jan  9 15:27:24 2012
From: nginx-forum at nginx.us (cn_nginxer)
Date: Mon, 09 Jan 2012 10:27:24 -0500
Subject: How to Windows auth working on nginx reverse proxy ???
In-Reply-To: <20120109140615.GE67687@mdounin.ru>
References: <20120109140615.GE67687@mdounin.ru>
Message-ID: <b2af2e980a8cbf1427aa68e2fe3c7ed5.NginxMailingListEnglish@forum.nginx.org>

Hello Maxim,

Many thanks for ur timely reply!

But I am still not quite understand about term "connection-based". I
checked with wireshark again, even NTLM auth schema, it did not just use
one socket to transport the whole data, it also uses several sockets to
transfer a HTML page in parallel. So it works just like Digest
Authentication, at least, in the auth process, they are the same. So I
am totally stuck in here, where is the session info and how does Win NT
keep them? And if what I have noticed was true, why Digest
Authentication worked but not NTLM?

A lonely Nginxer in china :-)

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,72871,220964#msg-220964


From nsabbi at officinedigitali.it  Mon Jan  9 16:20:27 2012
From: nsabbi at officinedigitali.it (nicola sabbi)
Date: Mon, 9 Jan 2012 17:20:27 +0100
Subject: PHP_ADMIN_VALUE and separate fcgi error_log - are they possible?
Message-ID: <CAKA77bw8Ly-r1CXW57Eu1KrX1ugMKATr5HdOxzqCL14ns5yn9A@mail.gmail.com>

Hello,
today I setup nginx 1.0.11 with multiple virtual hosts using Zend's CE
spawn-fcgi  as a backend.
All is working fine, but I can't
-propagate php's ADMIN values to the virtualhosts (I tried to add

location ~ \.php$

                {


                        fastcgi_param  PHP_ADMIN_VALUE
"open_basedir=/var/www/mysite
                        include        fastcgi_params;


                }

but a test php page doesn't see the correct value assigned to open_basedir)

-separate fcgi's error_log files, actually I couldn't even locate the file
where the logs are saved.

Can you help me to solve these problems, please?
Thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120109/0ad51a71/attachment.html>

From nginx-forum at nginx.us  Mon Jan  9 16:56:27 2012
From: nginx-forum at nginx.us (simonyarde)
Date: Mon, 09 Jan 2012 11:56:27 -0500
Subject: meaning of notice "using inherited sockets"
Message-ID: <8f140a3011ae14a6ffa64bf56b14a932.NginxMailingListEnglish@forum.nginx.org>

Can anyone confirm whether or not the below is cause for concern?

  2012/01/09 13:21:13 [notice] 364#0: using inherited sockets from
"/opt/nginx/   sbin/nginx"
  2012/01/09 13:21:13 [notice] 470#0: using inherited sockets from
"/opt/nginx/   sbin/nginx"

Ideally could someone point me to documentation for Nginx
warnings/notices.

S

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220968,220968#msg-220968


From nginx-forum at nginx.us  Mon Jan  9 17:01:37 2012
From: nginx-forum at nginx.us (starmonche)
Date: Mon, 09 Jan 2012 12:01:37 -0500
Subject: reverse proxy root folder only
In-Reply-To: <5170f7a3fb61c2f6067041772f82057a.NginxMailingListEnglish@forum.nginx.org>
References: <5170f7a3fb61c2f6067041772f82057a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <7f0562ab408111ea68fbbfdeec841677.NginxMailingListEnglish@forum.nginx.org>

That was exactly what I was looking for, thank you.

I have these locations defined in my server section:

location	~ ^/[^/]*$	{ proxy_pass http://myinternalsite.com; }
location	/subfolder	{ proxy_pass http://myinternalsite.com; }
location	/		{ return 444; }

So http://externalsite.com and http://externalsite.com/subfolder both
work but everything else fails.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220895,220969#msg-220969


From nginx-forum at nginx.us  Mon Jan  9 17:44:19 2012
From: nginx-forum at nginx.us (hirenshah.05)
Date: Mon, 09 Jan 2012 12:44:19 -0500
Subject: proxy_set_header X-Time (TIME)
Message-ID: <a09b5ac18ac072049e8de014f6e013f8.NginxMailingListEnglish@forum.nginx.org>

Hi,

I want add Current Time or Request Time in my header. 

Please let me know which variable I can use in,

add_header X-Time $(TIME)
proxy_set_header X-Time $(TIME)


Thanks,
Hiren

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220970,220970#msg-220970


From nginx-forum at nginx.us  Mon Jan  9 18:50:34 2012
From: nginx-forum at nginx.us (piotr)
Date: Mon, 09 Jan 2012 13:50:34 -0500
Subject: proxy_ignore_headers doesn't seem to work
In-Reply-To: <20120105031856.GM67687@mdounin.ru>
References: <20120105031856.GM67687@mdounin.ru>
Message-ID: <f4bac04c0a31ac56d0602c796133c94c.NginxMailingListEnglish@forum.nginx.org>

Maxim,

Yes, i've missed this "inactive" param - unfortunately expire time from
upstream server was the same as default inactive value, with was
misleading for me.

May I suggest to add additional info about this inactive param to
http://wiki.nginx.org/HttpProxyModule#proxy_cache_valid page?

Thank you for your help!

Regards,
p.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220058,220972#msg-220972


From nginx-forum at nginx.us  Mon Jan  9 19:00:46 2012
From: nginx-forum at nginx.us (locojohn)
Date: Mon, 09 Jan 2012 14:00:46 -0500
Subject: PHP_ADMIN_VALUE and separate fcgi error_log - are they possible?
In-Reply-To: <CAKA77bw8Ly-r1CXW57Eu1KrX1ugMKATr5HdOxzqCL14ns5yn9A@mail.gmail.com>
References: <CAKA77bw8Ly-r1CXW57Eu1KrX1ugMKATr5HdOxzqCL14ns5yn9A@mail.gmail.com>
Message-ID: <3b3e5ec1e5145c7a3d25f4d0825add13.NginxMailingListEnglish@forum.nginx.org>

Hi,

It is possible when "open_basedir" option is set in secure way that
disables setting it later , for example via PHP 5.3 featured host file
sections (http://php.net/manual/en/ini.sections.php), or in the
php-fpm.conf file via "php_admin_value[open_basedir]", etc.   Check and
come back to let us know.

Andrejs

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220966,220973#msg-220973


From lists at ruby-forum.com  Mon Jan  9 19:18:17 2012
From: lists at ruby-forum.com (dhaval d.)
Date: Mon, 09 Jan 2012 20:18:17 +0100
Subject: proxy_set_header X-Time (TIME)
In-Reply-To: <a09b5ac18ac072049e8de014f6e013f8.NginxMailingListEnglish@forum.nginx.org>
References: <a09b5ac18ac072049e8de014f6e013f8.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <10c2ada162805ea95914f8c0e82f5981@ruby-forum.com>

you can use $time_local in order to do that.

hirenshah.05 wrote in post #1040101:
> Hi,
>
> I want add Current Time or Request Time in my header.
>
> Please let me know which variable I can use in,
>
> add_header X-Time $(TIME)
> proxy_set_header X-Time $(TIME)
>
>
> Thanks,
> Hiren
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,220970,220970#msg-220970

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


From nginx-forum at nginx.us  Mon Jan  9 19:25:38 2012
From: nginx-forum at nginx.us (hirenshah.05)
Date: Mon, 09 Jan 2012 14:25:38 -0500
Subject: proxy_set_header X-Time (TIME)
In-Reply-To: <10c2ada162805ea95914f8c0e82f5981@ruby-forum.com>
References: <10c2ada162805ea95914f8c0e82f5981@ruby-forum.com>
Message-ID: <29d0e29abfec453b81c06b94dc34c516.NginxMailingListEnglish@forum.nginx.org>

Thanks for your reply.

I tried that. But looks like $time_local is only available for Log.

It does not recognize $time_local in Server section. 

May be I am missing something. Is there way to make these variable
available in anywhere in nginx conf ?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220970,220975#msg-220975


From brian at akins.org  Mon Jan  9 20:13:49 2012
From: brian at akins.org (Brian Akins)
Date: Mon, 9 Jan 2012 15:13:49 -0500
Subject: proxy_set_header X-Time (TIME)
In-Reply-To: <29d0e29abfec453b81c06b94dc34c516.NginxMailingListEnglish@forum.nginx.org>
References: <10c2ada162805ea95914f8c0e82f5981@ruby-forum.com>
	<29d0e29abfec453b81c06b94dc34c516.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CAO3qTJVo68K=GPQM=MJWiMkjHLt+wSby5HuM--Y3fj+Jvda0BQ@mail.gmail.com>

something like:

set_by_lua $my_time "return ngx.time()";
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120109/b72c8b7b/attachment-0001.html>

From lists at ruby-forum.com  Mon Jan  9 20:27:59 2012
From: lists at ruby-forum.com (dhaval d.)
Date: Mon, 09 Jan 2012 21:27:59 +0100
Subject: proxy_set_header X-Time (TIME)
In-Reply-To: <CAO3qTJVo68K=GPQM=MJWiMkjHLt+wSby5HuM--Y3fj+Jvda0BQ@mail.gmail.com>
References: <a09b5ac18ac072049e8de014f6e013f8.NginxMailingListEnglish@forum.nginx.org>
	<10c2ada162805ea95914f8c0e82f5981@ruby-forum.com>
	<29d0e29abfec453b81c06b94dc34c516.NginxMailingListEnglish@forum.nginx.org>
	<CAO3qTJVo68K=GPQM=MJWiMkjHLt+wSby5HuM--Y3fj+Jvda0BQ@mail.gmail.com>
Message-ID: <f7907b16802356383d3824520584bf58@ruby-forum.com>

yeah that is a good idea to use lua-nginx module or if you have built 
your nginx with perl then you can have a look at this as well

http://wiki.nginx.org/EmbeddedPerlModule

It is pretty much the same idea as Brian indicated.

Brian Akins wrote in post #1040119:
> something like:
>
> set_by_lua $my_time "return ngx.time()";

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


From nginx-forum at nginx.us  Mon Jan  9 21:00:47 2012
From: nginx-forum at nginx.us (hirenshah.05)
Date: Mon, 09 Jan 2012 16:00:47 -0500
Subject: proxy_set_header X-Time (TIME)
In-Reply-To: <a09b5ac18ac072049e8de014f6e013f8.NginxMailingListEnglish@forum.nginx.org>
References: <a09b5ac18ac072049e8de014f6e013f8.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <4b946c24929e2e51d0d34e2b8f2c34c0.NginxMailingListEnglish@forum.nginx.org>

Thanks Brian and Dhaval. 

So there is no way other than installing additional module ??

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220970,220979#msg-220979


From nginx-forum at nginx.us  Mon Jan  9 21:10:43 2012
From: nginx-forum at nginx.us (hirenshah.05)
Date: Mon, 09 Jan 2012 16:10:43 -0500
Subject: proxy_set_header X-Time (TIME)
In-Reply-To: <a09b5ac18ac072049e8de014f6e013f8.NginxMailingListEnglish@forum.nginx.org>
References: <a09b5ac18ac072049e8de014f6e013f8.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <a2cb627d63f50e5d8bfcf148fb786a1b.NginxMailingListEnglish@forum.nginx.org>

Thanks for your help.

I found this,
http://wiki.nginx.org/HttpSsiModule#.24date_gmt

I am using $date_gmt variable.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220970,220980#msg-220980


From emmanueloga at gmail.com  Mon Jan  9 21:46:44 2012
From: emmanueloga at gmail.com (Emmanuel Oga)
Date: Mon, 9 Jan 2012 18:46:44 -0300
Subject: setting +input+ headers for an upstream server depending on the
	request's path+method
Message-ID: <CAMsuM5qATz6YNo+2E6qTHkCeVPGw0KYYuLP0vYqKMxGqCqfU6A@mail.gmail.com>

Hello, I'm experimenting with having nginx prepare the request to my
application server in such way I can avoid implementing a router
there.

In this example, I tried to simulate rails' restful routes for a blog
post resource:

http://pastie.org/3156651

So if I hit an url like POST /posts, nginx should serve it through the
upstream app, passing the app parameters like:

  more_set_input_headers "X-NS: POSTS" "X-ACTION: create";

using headers-more module to set input headers. On my app server,
Instead of parsing the routes, I could just take a look on the headers
to know who should handle the response.

The configuration on that pastie works but is kind of verbose. Also,
I'm not sure how well it would work if I had tens or thousands of
routes (the idea would be to generate the nginx configuration from a
simpler manifest file, so you could define so called 'restfull routes'
without much effort).

Can you think of a better way to do this?

-- 
--------------------------------------------------------------
EmmanuelOga.com - Software Developer


From nginx-forum at nginx.us  Tue Jan 10 02:23:45 2012
From: nginx-forum at nginx.us (cn_nginxer)
Date: Mon, 09 Jan 2012 21:23:45 -0500
Subject: How to Windows auth working on nginx reverse proxy ???
In-Reply-To: <20120109140615.GE67687@mdounin.ru>
References: <20120109140615.GE67687@mdounin.ru>
Message-ID: <f2363d6fdab35dccb27881cb9bdd6b9b.NginxMailingListEnglish@forum.nginx.org>

Hello Nginx Team,

I finally get the idea about the connection.
It is not a TCP connection, it is a persude connection in HTTP level.
The NTLM will keep the authenticated HTTP connections info for
subsequent connections, so in this case, it should be a bundle TCP
authed connections. And for sure, if it works in this way, every proxy
won't work with this evil!!! Since this is not the first time I face
difficultis with MS system, I totally do not understand why MS could do
sth that compliable with industrial standard?! I should say MS makes
more valuable, I damned love it!

cn_nginxer

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,72871,220984#msg-220984


From ru at nginx.com  Tue Jan 10 07:17:49 2012
From: ru at nginx.com (Ruslan Ermilov)
Date: Tue, 10 Jan 2012 11:17:49 +0400
Subject: meaning of notice "using inherited sockets"
In-Reply-To: <8f140a3011ae14a6ffa64bf56b14a932.NginxMailingListEnglish@forum.nginx.org>
References: <8f140a3011ae14a6ffa64bf56b14a932.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120110071749.GA16010@lo0.su>

On Mon, Jan 09, 2012 at 11:56:27AM -0500, simonyarde wrote:
> Can anyone confirm whether or not the below is cause for concern?

Notices are normal (notification) messages.

>   2012/01/09 13:21:13 [notice] 364#0: using inherited sockets from
> "/opt/nginx/   sbin/nginx"
>   2012/01/09 13:21:13 [notice] 470#0: using inherited sockets from
> "/opt/nginx/   sbin/nginx"

In this case, an nginx binary is being replaced, as per
http://nginx.org/en/docs/control.html#upgrade
and a new binary gains control over listening sockets of the
old binary.

> Ideally could someone point me to documentation for Nginx
> warnings/notices.


From nginx-forum at nginx.us  Tue Jan 10 07:22:08 2012
From: nginx-forum at nginx.us (supportnti)
Date: Tue, 10 Jan 2012 02:22:08 -0500
Subject: what is ngx_cycle_t represent
In-Reply-To: <d97ba48596c42adca7ac39d360f52e87@none.at>
References: <d97ba48596c42adca7ac39d360f52e87@none.at>
Message-ID: <72d8f17f37249c1727006cb9e7f6ab61.NginxMailingListEnglish@forum.nginx.org>

Hi Aleks

Thanks for the response. Can you please give me a description of what
ngx_cycle_t and what does it represent? I mean is it a request-response
cycle for each request that comes in? Or is it something else?

Thanks

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220954,220986#msg-220986


From kofreestyler at gmail.com  Tue Jan 10 07:30:48 2012
From: kofreestyler at gmail.com (freestyler)
Date: Tue, 10 Jan 2012 15:30:48 +0800
Subject: what is ngx_cycle_t represent
In-Reply-To: <72d8f17f37249c1727006cb9e7f6ab61.NginxMailingListEnglish@forum.nginx.org>
References: <d97ba48596c42adca7ac39d360f52e87@none.at>
	<72d8f17f37249c1727006cb9e7f6ab61.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CA+uYZKt6S-RnD9Mcc5biTAZWDx3SgVpFiU6YYWF7t3ntttVkRg@mail.gmail.com>

I think it represents the process life cycle.


On Tue, Jan 10, 2012 at 3:22 PM, supportnti <nginx-forum at nginx.us> wrote:
> Hi Aleks
>
> Thanks for the response. Can you please give me a description of what
> ngx_cycle_t and what does it represent? I mean is it a request-response
> cycle for each request that comes in? Or is it something else?
>
> Thanks
>
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220954,220986#msg-220986
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From mdounin at mdounin.ru  Tue Jan 10 07:46:57 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 10 Jan 2012 11:46:57 +0400
Subject: what is ngx_cycle_t represent
In-Reply-To: <72d8f17f37249c1727006cb9e7f6ab61.NginxMailingListEnglish@forum.nginx.org>
References: <d97ba48596c42adca7ac39d360f52e87@none.at>
	<72d8f17f37249c1727006cb9e7f6ab61.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120110074656.GG67687@mdounin.ru>

Hello!

On Tue, Jan 10, 2012 at 02:22:08AM -0500, supportnti wrote:

> Hi Aleks
> 
> Thanks for the response. Can you please give me a description of what
> ngx_cycle_t and what does it represent? I mean is it a request-response
> cycle for each request that comes in? Or is it something else?

It represents global data and configuration of a event loop (that 
is, whole nginx).  New cycle is created on a reconfiguration.

Maxim Dounin


From lists at ruby-forum.com  Tue Jan 10 08:42:45 2012
From: lists at ruby-forum.com (Maik Unger)
Date: Tue, 10 Jan 2012 09:42:45 +0100
Subject: nGinX HTTP-Code 200 / 404
Message-ID: <cb59c6bf53e47e03307b54a298f52726@ruby-forum.com>

Hy,

We have that problem, that a site is not reachable throught nginx with
the domain name (for example www.foobar.org), even the websites are
reachable without nginx directly to apache.

This problem is not time specific, so that we have no solution at this
moment.

What is the problem here?

Best Regards,


Maik Unger

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


From nginx-forum at nginx.us  Tue Jan 10 09:13:02 2012
From: nginx-forum at nginx.us (supportnti)
Date: Tue, 10 Jan 2012 04:13:02 -0500
Subject: what is ngx_cycle_t represent
In-Reply-To: <20120110074656.GG67687@mdounin.ru>
References: <20120110074656.GG67687@mdounin.ru>
Message-ID: <434a0c74184cf6ca6e68cc7bfefec218.NginxMailingListEnglish@forum.nginx.org>

Hi Maxim,

Thanks for the response. So ngx_cycle_t represents global data and
configuration per event loop. What does ngx_conf_t represent? Is it
related to ngx_conf_t? If so, how?

Thanks

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220954,220993#msg-220993


From nginx-forum at nginx.us  Tue Jan 10 09:13:24 2012
From: nginx-forum at nginx.us (supportnti)
Date: Tue, 10 Jan 2012 04:13:24 -0500
Subject: what is ngx_cycle_t represent
In-Reply-To: <434a0c74184cf6ca6e68cc7bfefec218.NginxMailingListEnglish@forum.nginx.org>
References: <20120110074656.GG67687@mdounin.ru>
	<434a0c74184cf6ca6e68cc7bfefec218.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <2467c42d1bb569e6f5b4cb6d6b49db15.NginxMailingListEnglish@forum.nginx.org>

Hi Maxim,

Thanks for the response. So ngx_cycle_t represents global data and
configuration per event loop. What does ngx_conf_t represent? Is it
related to ngx_cycle_t? If so, how?

Thanks

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220954,220994#msg-220994


From andrew at nginx.com  Tue Jan 10 09:31:52 2012
From: andrew at nginx.com (Andrew Alexeev)
Date: Tue, 10 Jan 2012 13:31:52 +0400
Subject: meaning of notice "using inherited sockets"
In-Reply-To: <8f140a3011ae14a6ffa64bf56b14a932.NginxMailingListEnglish@forum.nginx.org>
References: <8f140a3011ae14a6ffa64bf56b14a932.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <A07901CB-D019-491D-A495-A125CC95F802@nginx.com>

On Jan 9, 2012, at 8:56 PM, simonyarde wrote:

> Can anyone confirm whether or not the below is cause for concern?
> 
>  2012/01/09 13:21:13 [notice] 364#0: using inherited sockets from
> "/opt/nginx/   sbin/nginx"
>  2012/01/09 13:21:13 [notice] 470#0: using inherited sockets from
> "/opt/nginx/   sbin/nginx"
> 
> Ideally could someone point me to documentation for Nginx
> warnings/notices.

Such documentation is planned to appear soon as part of our efforts on making the documentation complete.


From mdounin at mdounin.ru  Tue Jan 10 10:25:04 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 10 Jan 2012 14:25:04 +0400
Subject: nGinX HTTP-Code 200 / 404
In-Reply-To: <cb59c6bf53e47e03307b54a298f52726@ruby-forum.com>
References: <cb59c6bf53e47e03307b54a298f52726@ruby-forum.com>
Message-ID: <20120110102504.GK67687@mdounin.ru>

Hello!

On Tue, Jan 10, 2012 at 09:42:45AM +0100, Maik Unger wrote:

> Hy,
> 
> We have that problem, that a site is not reachable throught nginx with
> the domain name (for example www.foobar.org), even the websites are
> reachable without nginx directly to apache.
> 
> This problem is not time specific, so that we have no solution at this
> moment.
> 
> What is the problem here?

Most likely you are using wrong config.

By default nginx will use hostname as specified in proxy_pass 
directive, i.e.

    proxy_pass http://127.0.0.1:8080;

will generate request with Host header set to "127.0.0.1:8080", 
not to "www.foobar.org".  Your backend server won't recognize it 
if it's configured to serve many virtual hosts, and 
"www.foobar.org" is just one of them.

To preserve the hostname from the original request (as sent to 
nginx server), use 

    proxy_pass http://127.0.0.1:8080;
    proxy_set_header Host $host;

Maxim Dounin


From lists at ruby-forum.com  Tue Jan 10 10:34:59 2012
From: lists at ruby-forum.com (Maik Unger)
Date: Tue, 10 Jan 2012 11:34:59 +0100
Subject: nGinX HTTP-Code 200 / 404
In-Reply-To: <20120110102504.GK67687@mdounin.ru>
References: <cb59c6bf53e47e03307b54a298f52726@ruby-forum.com>
	<20120110102504.GK67687@mdounin.ru>
Message-ID: <46f0616c5fe0385e7bceebaf4db1fb3f@ruby-forum.com>

Hello,

Out Config-File is:

server {
        listen                          xxx.xxx.xxx.xxx:80;
        server_name                     domain.foobar.org;
        access_log 
/var/www/html/domain.foobar.org/logs/nginx_domain.foobar.org_access.log 
main;

        location / {
                proxy_pass 
http://xxx.xxx.xxx.xxx:8080;
                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;
                proxy_set_header                Referer 
$http_referer;
                proxy_set_header                Accept-Language 
$http_accept_language;

                client_max_body_size            64m;
                client_body_buffer_size         128k;

                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;

                proxy_cache                     STATIC;
                proxy_cache_valid               200 30m;
                proxy_cache_valid               any 1m;
                proxy_cache_use_stale           error timeout 
invalid_header updating http_500 http_502 http_503 http_504;
        }
}

As you can see, we have already set the parameter proxy_set_header. The 
problem occurs randomly, not all the time. When the problem occurs, we 
got the apache server startpage with the servers hostname.

Maxim Dounin wrote in post #1040185:
> Hello!
>
> On Tue, Jan 10, 2012 at 09:42:45AM +0100, Maik Unger wrote:
>
>> Hy,
>>
>> We have that problem, that a site is not reachable throught nginx with
>> the domain name (for example www.foobar.org), even the websites are
>> reachable without nginx directly to apache.
>>
>> This problem is not time specific, so that we have no solution at this
>> moment.
>>
>> What is the problem here?
>
> Most likely you are using wrong config.
>
> By default nginx will use hostname as specified in proxy_pass
> directive, i.e.
>
>     proxy_pass http://127.0.0.1:8080;
>
> will generate request with Host header set to "127.0.0.1:8080",
> not to "www.foobar.org".  Your backend server won't recognize it
> if it's configured to serve many virtual hosts, and
> "www.foobar.org" is just one of them.
>
> To preserve the hostname from the original request (as sent to
> nginx server), use
>
>     proxy_pass http://127.0.0.1:8080;
>     proxy_set_header Host $host;
>
> Maxim Dounin

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


From mdounin at mdounin.ru  Tue Jan 10 10:43:36 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 10 Jan 2012 14:43:36 +0400
Subject: what is ngx_cycle_t represent
In-Reply-To: <2467c42d1bb569e6f5b4cb6d6b49db15.NginxMailingListEnglish@forum.nginx.org>
References: <20120110074656.GG67687@mdounin.ru>
	<434a0c74184cf6ca6e68cc7bfefec218.NginxMailingListEnglish@forum.nginx.org>
	<2467c42d1bb569e6f5b4cb6d6b49db15.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120110104336.GM67687@mdounin.ru>

Hello!

On Tue, Jan 10, 2012 at 04:13:24AM -0500, supportnti wrote:

> Hi Maxim,
> 
> Thanks for the response. So ngx_cycle_t represents global data and
> configuration per event loop. What does ngx_conf_t represent? Is it
> related to ngx_cycle_t? If so, how?

The ngx_conf_t is a structure used during configuration parsing.  
It holds various related data, e.g. pointer to a new cycle 
structure, current command arguments and so on.

Use the Source, Luke!

Maxim Dounin


From jelledejong at powercraft.nl  Tue Jan 10 11:21:15 2012
From: jelledejong at powercraft.nl (Jelle de Jong)
Date: Tue, 10 Jan 2012 12:21:15 +0100
Subject: https proxy to apache server
Message-ID: <4F0C1F2B.9010902@powercraft.nl>

I got an namebased vhost apache server running on 192.168.24.66 for
calendar.powercraft.nl on both port 80 and 443.

I need nginx to make a proxy pass for port 443 to the apache server. I
don't mind if it uses the port 80 or 443 with ssl to proxy the data.

I tried the bellow and some variations based on some docs. and wiki
pages I found, but I cant get it to work for port 443

server {
    listen      443;
    server_name calendar.powercraft.nl;
    ssl         on;
    ssl_certificate      /etc/ssl/certs/yessica-certificate.pem;
    ssl_certificate_key  /etc/ssl/certs/yessica-privatekey.pem;
    access_log  /var/log/nginx/calendar.secure.log;
    error_log   /var/log/nginx/calendar.error-secure.log;
    location / {
        #proxy_set_header   Host $host;
        proxy_set_header    X-Real-IP $remote_addr;
        proxy_set_header    X-Forwarded-For $proxy_add_x_forwarded_for;
        proxy_redirect      off;
        proxy_pass          http://192.168.24.66/;
    }
}

https://calendar.powercraft.nl/ is giving me SSL received a record that
exceeded the maximum permissible length. (Error code:
ssl_error_rx_record_too_long)

Does somebody know what I am doing wrong and may provide a working
example to make a https proxy with nginx to an apache server?

I tried searching the mailing-list archive for the last year but didn't
find a similar question. (my apologies if I missed something)

Thanks in advance,

Kind regards,

Jelle de Jong


From brian at akins.org  Tue Jan 10 11:56:00 2012
From: brian at akins.org (Brian Akins)
Date: Tue, 10 Jan 2012 06:56:00 -0500
Subject: https proxy to apache server
In-Reply-To: <4F0C1F2B.9010902@powercraft.nl>
References: <4F0C1F2B.9010902@powercraft.nl>
Message-ID: <0609FA4C-A892-409E-A0A4-31BEE78C486E@akins.org>


AFAIK, you can't proxy directly to https.  Why are you trying to do this. Generally you terminate SSL on nginx the proxy to something else in cases like this.

If you really (for some reason) needed to proxy over https, you could put something like stunnel (or stud) and have nginx talk to it.



From jelledejong at powercraft.nl  Tue Jan 10 12:12:16 2012
From: jelledejong at powercraft.nl (Jelle de Jong)
Date: Tue, 10 Jan 2012 13:12:16 +0100
Subject: https proxy to apache server
In-Reply-To: <0609FA4C-A892-409E-A0A4-31BEE78C486E@akins.org>
References: <4F0C1F2B.9010902@powercraft.nl>
	<0609FA4C-A892-409E-A0A4-31BEE78C486E@akins.org>
Message-ID: <4F0C2B20.6050506@powercraft.nl>

On 10/01/12 12:56, Brian Akins wrote:
> AFAIK, you can't proxy directly to https.  Why are you trying to do
> this. Generally you terminate SSL on nginx the proxy to something
> else in cases like this. If you really (for some reason) needed to
> proxy over https, you could put something like stunnel (or stud) and
> have nginx talk to it.

I don?t mind proxing with nginx port 443 to apache port 80, as long as
the connection between nginx and the client is using https.

I want this because I got only one IPv4 address and multiple local
servers with services running on port 80 or 443. Proxing for nginx port
80 to apache port 80 works perfectly. I now want to have an https
connection also.

Thanks in advance,

Kind regards,

Jelle de Jong


From nsabbi at officinedigitali.it  Tue Jan 10 12:14:00 2012
From: nsabbi at officinedigitali.it (nicola sabbi)
Date: Tue, 10 Jan 2012 13:14:00 +0100
Subject: PHP_ADMIN_VALUE and separate fcgi error_log - are they possible?
In-Reply-To: <3b3e5ec1e5145c7a3d25f4d0825add13.NginxMailingListEnglish@forum.nginx.org>
References: <CAKA77bw8Ly-r1CXW57Eu1KrX1ugMKATr5HdOxzqCL14ns5yn9A@mail.gmail.com>
	<3b3e5ec1e5145c7a3d25f4d0825add13.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CAKA77bwduyfajW8SwGOs9+fnTRqV3RGnX_rkDm=rZ8c0bQf+ZQ@mail.gmail.com>

2012/1/9 locojohn <nginx-forum at nginx.us>

> Hi,
>
> It is possible when "open_basedir" option is set in secure way that
> disables setting it later , for example via PHP 5.3 featured host file
> sections (http://php.net/manual/en/ini.sections.php), or in the
> php-fpm.conf file via "php_admin_value[open_basedir]", etc.   Check and
> come back to let us know.
>
> Andrejs
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,220966,220973#msg-220973
>
>
I've just installed php-fpm. open_basedir isn't set anywhere and the row

fastcgi_param  PHP_ADMIN_VALUE "open_basedir=/mysite";

(nor PHP_VALUE) doesn't change anything.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120110/91227a9d/attachment.html>

From nginx-forum at nginx.us  Tue Jan 10 12:25:02 2012
From: nginx-forum at nginx.us (locojohn)
Date: Tue, 10 Jan 2012 07:25:02 -0500
Subject: PHP_ADMIN_VALUE and separate fcgi error_log - are they possible?
In-Reply-To: <CAKA77bwduyfajW8SwGOs9+fnTRqV3RGnX_rkDm=rZ8c0bQf+ZQ@mail.gmail.com>
References: <CAKA77bwduyfajW8SwGOs9+fnTRqV3RGnX_rkDm=rZ8c0bQf+ZQ@mail.gmail.com>
Message-ID: <c3023f4b285d983183604c9a3dd67ba5.NginxMailingListEnglish@forum.nginx.org>

OK, so phpinfo() in your virtual host doesn't output "/mysite" as a
value for open_basedir?

What does php-fpm.conf look like?   

What does php.ini look like?

Andrew

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220966,221005#msg-221005


From nginx-forum at nginx.us  Tue Jan 10 12:38:04 2012
From: nginx-forum at nginx.us (simonyarde)
Date: Tue, 10 Jan 2012 07:38:04 -0500
Subject: meaning of notice "using inherited sockets"
In-Reply-To: <A07901CB-D019-491D-A495-A125CC95F802@nginx.com>
References: <A07901CB-D019-491D-A495-A125CC95F802@nginx.com>
Message-ID: <35603f28f4604c88bddaef03da8cb854.NginxMailingListEnglish@forum.nginx.org>

Thanks that's great.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220968,221006#msg-221006


From mdounin at mdounin.ru  Tue Jan 10 13:04:44 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 10 Jan 2012 17:04:44 +0400
Subject: nGinX HTTP-Code 200 / 404
In-Reply-To: <46f0616c5fe0385e7bceebaf4db1fb3f@ruby-forum.com>
References: <cb59c6bf53e47e03307b54a298f52726@ruby-forum.com>
	<20120110102504.GK67687@mdounin.ru>
	<46f0616c5fe0385e7bceebaf4db1fb3f@ruby-forum.com>
Message-ID: <20120110130444.GN67687@mdounin.ru>

Hello!

On Tue, Jan 10, 2012 at 11:34:59AM +0100, Maik Unger wrote:

> Hello,
> 
> Out Config-File is:
> 
> server {
>         listen                          xxx.xxx.xxx.xxx:80;
>         server_name                     domain.foobar.org;
>         access_log 
> /var/www/html/domain.foobar.org/logs/nginx_domain.foobar.org_access.log 
> main;
> 
>         location / {
>                 proxy_pass 
> http://xxx.xxx.xxx.xxx:8080;
>                 proxy_redirect                  off;
> 
>                 proxy_set_header                Host            $host;

[...]

>                 proxy_cache                     STATIC;
>                 proxy_cache_valid               200 30m;
>                 proxy_cache_valid               any 1m;
>                 proxy_cache_use_stale           error timeout 
> invalid_header updating http_500 http_502 http_503 http_504;
>         }
> }
> 
> As you can see, we have already set the parameter proxy_set_header. The 
> problem occurs randomly, not all the time. When the problem occurs, we 
> got the apache server startpage with the servers hostname.

Which server's hostname?  The one matching "domain.foobar.org", or 
something else?

Two possible causes I would consider are:

1. Cache is poisoned by unrelated requests.  This may happen if 
you use the same proxy_cache in other server{} blocks, or if the 
server{} block in question may be used for other hostnames as 
well (i.e. it's default one).

2. Something wrong happens on backend and it returns wrong 
response(s) sometimes.  This may happen for a second or even 
a single request, but as you have cache configured - the response is 
cached for a relatively long time and hence you see it.  E.g. this 
may be caused by some periodic backend's config regeneration.

Maxim Dounin



> 
> Maxim Dounin wrote in post #1040185:
> > Hello!
> >
> > On Tue, Jan 10, 2012 at 09:42:45AM +0100, Maik Unger wrote:
> >
> >> Hy,
> >>
> >> We have that problem, that a site is not reachable throught nginx with
> >> the domain name (for example www.foobar.org), even the websites are
> >> reachable without nginx directly to apache.
> >>
> >> This problem is not time specific, so that we have no solution at this
> >> moment.
> >>
> >> What is the problem here?
> >
> > Most likely you are using wrong config.
> >
> > By default nginx will use hostname as specified in proxy_pass
> > directive, i.e.
> >
> >     proxy_pass http://127.0.0.1:8080;
> >
> > will generate request with Host header set to "127.0.0.1:8080",
> > not to "www.foobar.org".  Your backend server won't recognize it
> > if it's configured to serve many virtual hosts, and
> > "www.foobar.org" is just one of them.
> >
> > To preserve the hostname from the original request (as sent to
> > nginx server), use
> >
> >     proxy_pass http://127.0.0.1:8080;
> >     proxy_set_header Host $host;
> >
> > Maxim Dounin
> 
> -- 
> Posted via http://www.ruby-forum.com/.
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From mdounin at mdounin.ru  Tue Jan 10 14:38:50 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 10 Jan 2012 18:38:50 +0400
Subject: https proxy to apache server
In-Reply-To: <4F0C1F2B.9010902@powercraft.nl>
References: <4F0C1F2B.9010902@powercraft.nl>
Message-ID: <20120110143850.GO67687@mdounin.ru>

Hello!

On Tue, Jan 10, 2012 at 12:21:15PM +0100, Jelle de Jong wrote:

> I got an namebased vhost apache server running on 192.168.24.66 for
> calendar.powercraft.nl on both port 80 and 443.
> 
> I need nginx to make a proxy pass for port 443 to the apache server. I
> don't mind if it uses the port 80 or 443 with ssl to proxy the data.
> 
> I tried the bellow and some variations based on some docs. and wiki
> pages I found, but I cant get it to work for port 443
> 
> server {
>     listen      443;
>     server_name calendar.powercraft.nl;
>     ssl         on;
>     ssl_certificate      /etc/ssl/certs/yessica-certificate.pem;
>     ssl_certificate_key  /etc/ssl/certs/yessica-privatekey.pem;
>     access_log  /var/log/nginx/calendar.secure.log;
>     error_log   /var/log/nginx/calendar.error-secure.log;
>     location / {
>         #proxy_set_header   Host $host;
>         proxy_set_header    X-Real-IP $remote_addr;
>         proxy_set_header    X-Forwarded-For $proxy_add_x_forwarded_for;
>         proxy_redirect      off;
>         proxy_pass          http://192.168.24.66/;

Just use "https" instead of "http" here, i.e.

          proxy_pass          https://192.168.24.66/;

>     }
> }
> 
> https://calendar.powercraft.nl/ is giving me SSL received a record that
> exceeded the maximum permissible length. (Error code:
> ssl_error_rx_record_too_long)
> 
> Does somebody know what I am doing wrong and may provide a working
> example to make a https proxy with nginx to an apache server?
> 
> I tried searching the mailing-list archive for the last year but didn't
> find a similar question. (my apologies if I missed something)

Maxim Dounin


From nsabbi at officinedigitali.it  Tue Jan 10 14:48:15 2012
From: nsabbi at officinedigitali.it (nicola sabbi)
Date: Tue, 10 Jan 2012 15:48:15 +0100
Subject: PHP_ADMIN_VALUE and separate fcgi error_log - are they possible?
In-Reply-To: <c3023f4b285d983183604c9a3dd67ba5.NginxMailingListEnglish@forum.nginx.org>
References: <CAKA77bwduyfajW8SwGOs9+fnTRqV3RGnX_rkDm=rZ8c0bQf+ZQ@mail.gmail.com>
	<c3023f4b285d983183604c9a3dd67ba5.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CAKA77by3nLTjN7jvM0r0ZPGADTyKRHKaieceuzZqgJYggaS0QA@mail.gmail.com>

2012/1/10 locojohn <nginx-forum at nginx.us>

> OK, so phpinfo() in your virtual host doesn't output "/mysite" as a
> value for open_basedir?
>

argh, now it's working using PHP_VALUE rather than PHP_ADMIN_VALUE.


>
> What does php-fpm.conf look like?
>
> What does php.ini look like?
>
> Andrew
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,220966,221005#msg-221005
>
>
 What about having separate log files, especially separate error_log?
Thanks for all your help.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120110/c8dfff78/attachment.html>

From jelledejong at powercraft.nl  Tue Jan 10 15:17:21 2012
From: jelledejong at powercraft.nl (Jelle de Jong)
Date: Tue, 10 Jan 2012 16:17:21 +0100
Subject: https proxy to apache server
In-Reply-To: <20120110143850.GO67687@mdounin.ru>
References: <4F0C1F2B.9010902@powercraft.nl>
	<20120110143850.GO67687@mdounin.ru>
Message-ID: <4F0C5681.5090201@powercraft.nl>

On 10/01/12 15:38, Maxim Dounin wrote:
> Just use "https" instead of "http" here, i.e.
>           proxy_pass          https://192.168.24.66/;

I tried that, (did it again to be sure). I think the issues may be
somewhere else...

I removed the port 80 part of the config and only have the 443 now.

# checking: nginx is listing
# lsof -i :443
COMMAND   PID     USER   FD   TYPE DEVICE SIZE/OFF NODE NAME
nginx   30136     root   20u  IPv4 482955      0t0  TCP *:https (LISTEN)
nginx   30137 www-data   20u  IPv4 482955      0t0  TCP *:https (LISTEN)
nginx   30138 www-data   20u  IPv4 482955      0t0  TCP *:https (LISTEN)
nginx   30139 www-data   20u  IPv4 482955      0t0  TCP *:https (LISTEN)
nginx   30140 www-data   20u  IPv4 482955      0t0  TCP *:https (LISTEN)

# checking: wierd response from ssl check, what is wrong?
# openssl s_client -connect localhost:443
CONNECTED(00000003)
31082:error:140770FC:SSL routines:SSL23_GET_SERVER_HELLO:unknown
protocol:s23_clnt.c:607:

# nginx should be able to read the certs?
# output shorted for mail purpose
# su -c 'ls -hal /etc/ssl/certs/yessica-*' www-data
-rw-r--r-- 1 root root 1.4K Jan  7 16:42 /etc/ssl/certs/cert-request.pem
-rw-r--r-- 1 root root 2.9K Jan  7 16:42 /etc/ssl/certs/cert.pem
-r-------- 1 root root 1.7K Jan  7 16:42 /etc/ssl/certs/privatekey.pem

# the config...
server {
    listen      443;
    server_name calendar.powercraft.nl;
    ssl         on;
    ssl_certificate     /etc/ssl/certs/yessica-certificate.pem;
    ssl_certificate_key /etc/ssl/certs/yessica--privatekey.pem;
    access_log          /var/log/nginx/calendar.secure.log;
    error_log           /var/log/nginx/calendar.error-secure.log;
    location / {
        proxy_set_header    Host $host;
        proxy_set_header    X-Real-IP $remote_addr;
        proxy_set_header    X-Forwarded-For $proxy_add_x_forwarded_for;
        proxy_redirect      off;
        proxy_pass          https://192.168.24.66/;
    }
}

Thanks in advance,

Kind regards,

Jelle de Jong


From jelledejong at powercraft.nl  Tue Jan 10 16:13:02 2012
From: jelledejong at powercraft.nl (Jelle de Jong)
Date: Tue, 10 Jan 2012 17:13:02 +0100
Subject: https proxy to apache server
In-Reply-To: <4F0C5681.5090201@powercraft.nl>
References: <4F0C1F2B.9010902@powercraft.nl>
	<20120110143850.GO67687@mdounin.ru>
	<4F0C5681.5090201@powercraft.nl>
Message-ID: <4F0C638E.1080007@powercraft.nl>

On 10/01/12 16:17, Jelle de Jong wrote:
> On 10/01/12 15:38, Maxim Dounin wrote:
>> Just use "https" instead of "http" here, i.e.
>>           proxy_pass          https://192.168.24.66/;
> 
> I tried that, (did it again to be sure). I think the issues may be
> somewhere else...

I had another server section with:

server {
    listen      80;
    listen      443;
    server_name guadec.powercraft.nl;
...

I removed the listen 443; line and everything started working....

I have no idea why this was interfering with the other server section
but it did?

I am OK for now :D

Thanks in advance,

Kind regards,

Jelle de Jong


From nginx-forum at nginx.us  Tue Jan 10 16:28:02 2012
From: nginx-forum at nginx.us (locojohn)
Date: Tue, 10 Jan 2012 11:28:02 -0500
Subject: PHP_ADMIN_VALUE and separate fcgi error_log - are they possible?
In-Reply-To: <CAKA77by3nLTjN7jvM0r0ZPGADTyKRHKaieceuzZqgJYggaS0QA@mail.gmail.com>
References: <CAKA77by3nLTjN7jvM0r0ZPGADTyKRHKaieceuzZqgJYggaS0QA@mail.gmail.com>
Message-ID: <b70630c654e70bc4772b4c21cb23ba40.NginxMailingListEnglish@forum.nginx.org>

Well, error_log and access_log directives let you specify separate error
and access logs per virtual host.  

Andrejs

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220966,221015#msg-221015


From brian at akins.org  Tue Jan 10 18:55:59 2012
From: brian at akins.org (Brian Akins)
Date: Tue, 10 Jan 2012 13:55:59 -0500
Subject: https proxy to apache server
In-Reply-To: <20120110143850.GO67687@mdounin.ru>
References: <4F0C1F2B.9010902@powercraft.nl>
	<20120110143850.GO67687@mdounin.ru>
Message-ID: <CAO3qTJVFZU9eQnWoQ8enCcwnhRRS8KPEyMHqQK_6yAnTa78-9Q@mail.gmail.com>

On Tue, Jan 10, 2012 at 9:38 AM, Maxim Dounin <mdounin at mdounin.ru> wrote:

>
> Just use "https" instead of "http" here, i.e.
>
>          proxy_pass          https://192.168.24.66/;
>
>

I learn something new everyday. I didn't know this was possible.

--Brian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120110/47f5728b/attachment.html>

From nginx-forum at nginx.us  Wed Jan 11 03:17:44 2012
From: nginx-forum at nginx.us (zzor123)
Date: Tue, 10 Jan 2012 22:17:44 -0500
Subject: visual studio 2010 crash error
Message-ID: <31bb9c0dd76fa36193432587c0aed556.NginxMailingListEnglish@forum.nginx.org>

nginx 1.1.12 + openssl 1.0.0 + zlib 1.2.5 + pcre 8.21

visual studio 2010 building..
nginx -s stop crash error..


openssl 0.9.8 crash no..

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221018,221018#msg-221018


From nginx-forum at nginx.us  Wed Jan 11 03:39:52 2012
From: nginx-forum at nginx.us (Florante)
Date: Tue, 10 Jan 2012 22:39:52 -0500
Subject: Mobile browser detection at server level
In-Reply-To: <28646c8afe0455c277b2e3a8a7b0c004.squirrel@www.digitalhit.com>
References: <28646c8afe0455c277b2e3a8a7b0c004.squirrel@www.digitalhit.com>
Message-ID: <6cafb5626a97547d33e49dc814a95bd1.NginxMailingListEnglish@forum.nginx.org>

Handset Detection has a good solution when it comes to detection and
redirection. There is also a new detection system coming out that allows
you to run the detection either from your server or HD's.  API kits are
currently available for PHP, Ruby, Coldfusion, ASP.net, Java and Python.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,183827,221019#msg-221019


From brian at akins.org  Wed Jan 11 03:55:11 2012
From: brian at akins.org (Brian Akins)
Date: Tue, 10 Jan 2012 22:55:11 -0500
Subject: Mobile browser detection at server level
In-Reply-To: <6cafb5626a97547d33e49dc814a95bd1.NginxMailingListEnglish@forum.nginx.org>
References: <28646c8afe0455c277b2e3a8a7b0c004.squirrel@www.digitalhit.com>
	<6cafb5626a97547d33e49dc814a95bd1.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CAO3qTJXwBZDj_3gy-xdo=_=TEvNHKS5o-qOJ150UoWkD+bkuOg@mail.gmail.com>

We found most of the "packaged" detection libraries/databases very, very
slow. It's hard to beat the performance of Lua embedded in nginx.

--Brian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120110/12edea6a/attachment.html>

From mdounin at mdounin.ru  Wed Jan 11 10:22:19 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Wed, 11 Jan 2012 14:22:19 +0400
Subject: visual studio 2010 crash error
In-Reply-To: <31bb9c0dd76fa36193432587c0aed556.NginxMailingListEnglish@forum.nginx.org>
References: <31bb9c0dd76fa36193432587c0aed556.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120111102218.GS67687@mdounin.ru>

Hello!

On Tue, Jan 10, 2012 at 10:17:44PM -0500, zzor123 wrote:

> nginx 1.1.12 + openssl 1.0.0 + zlib 1.2.5 + pcre 8.21
> 
> visual studio 2010 building..
> nginx -s stop crash error..
> 
> 
> openssl 0.9.8 crash no..

Which openssl version are you using, exactly?  There were problems 
under windows with openssl 1.0.0, though they seems to be fixed in 
more recent releases.  OpenSSL 1.0.0f should be ok.

Maxim Dounin


From defan at nginx.com  Wed Jan 11 13:17:32 2012
From: defan at nginx.com (Andrey Belov)
Date: Wed, 11 Jan 2012 17:17:32 +0400
Subject: mod_aclr2
Message-ID: <ED8F5676-8A10-4F60-94B8-C09812A4A954@nginx.com>

Hello!

Ever wondered how to set up NGINX as a local proxy for Apache 2.x with minimum efforts?

Check out this module:
https://github.com/defanator/mod_aclr2

Feedback is welcome!

--
Andrey Belov
http://nginx.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120111/3941374e/attachment.html>

From frumentius at gmail.com  Wed Jan 11 13:40:00 2012
From: frumentius at gmail.com (Joe)
Date: Wed, 11 Jan 2012 20:40:00 +0700
Subject: mod_aclr2
In-Reply-To: <ED8F5676-8A10-4F60-94B8-C09812A4A954@nginx.com>
References: <ED8F5676-8A10-4F60-94B8-C09812A4A954@nginx.com>
Message-ID: <CAPihCJVFh9o5s7s-_BejCU85w1pVS1iJLHuVzY6e5bQG2-E3Vw@mail.gmail.com>

Thank You Andrey.


Regards,
Joe


On Wed, Jan 11, 2012 at 8:17 PM, Andrey Belov <defan at nginx.com> wrote:

> Hello!
>
> Ever wondered how to set up NGINX as a local proxy for Apache 2.x with
> minimum efforts?
>
> Check out this module:
> https://github.com/defanator/mod_aclr2
>
> Feedback is welcome!
>
> --
> Andrey Belov
> http://nginx.com
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120111/7f935d74/attachment.html>

From appa at perusio.net  Wed Jan 11 13:54:39 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Wed, 11 Jan 2012 13:54:39 +0000
Subject: mod_aclr2
In-Reply-To: <ED8F5676-8A10-4F60-94B8-C09812A4A954@nginx.com>
References: <ED8F5676-8A10-4F60-94B8-C09812A4A954@nginx.com>
Message-ID: <87sjjmxs2o.wl%appa@perusio.net>

On 11 Jan 2012 13h17 WET, defan at nginx.com wrote:

> Hello!
>
> Ever wondered how to set up NGINX as a local proxy for Apache 2.x
> with minimum efforts?

Nice work Andrey. Opens up Nginx as a possibility for shared hosting
(keep using .htaccess) and voids any resistance people might have trying
Nginx out. It's sooo easy now :)

Thanks,
--- appa


From agentzh at gmail.com  Wed Jan 11 15:11:05 2012
From: agentzh at gmail.com (agentzh)
Date: Wed, 11 Jan 2012 23:11:05 +0800
Subject: [ANN] ngx_lua module v0.4.0 released!
In-Reply-To: <CAB4Tn6MGdGX1ZYXr99W1TWgkC6oZQLyq8Qv-XESXQzrZuUhM0g@mail.gmail.com>
References: <CAB4Tn6MGdGX1ZYXr99W1TWgkC6oZQLyq8Qv-XESXQzrZuUhM0g@mail.gmail.com>
Message-ID: <CAB4Tn6M9uDZ8j0jb24fiwj6VQ4GTaXgmeNHYXn=2=41TQ2N-Zw@mail.gmail.com>

Hi, folks!

After four months' active development of our contributors, I'm happy
to announce the v0.4.0 release of our ngx_lua module. You can get the
release tarball from the download page:

? ?https://github.com/chaoslawful/lua-nginx-module/tags

There's many many new features and bug fixes that you can find out in
the following complete change log for this release (it is just too
long to put here :P):

? ?http://wiki.nginx.org/HttpLuaModule#v0.4.0

Special thanks go to all of our contributors and users!

This module embeds the Lua 5.1 interpreter or LuaJIT 2.0 into the
nginx core and integrates the powerful Lua threads (aka Lua
coroutines) into the nginx event model by means of nginx subrequests.

Unlike Apache's mod_lua and Lighttpd's mod_magnet, Lua code written
atop this module can be 100% non-blocking on network traffic as long
as you use the ngx.location.capture or ngx.location.capture_multi
interfaces to let the Nginx core do all your requests to mysql,
postgresql, memcached, redis, upstream http web services, and etc etc
etc.

This module is also included and enabled by default in our
ngx_openresty bundle: http://openresty.org/

You can find the complete documentation for this module on the
following wiki page:

? ?http://wiki.nginx.org/HttpLuaModule

And you always get the latest source code from the git repository here:

? ?https://github.com/chaoslawful/lua-nginx-module

Have fun!
-agentzh


From smallfish.xy at gmail.com  Wed Jan 11 15:13:42 2012
From: smallfish.xy at gmail.com (smallfish)
Date: Wed, 11 Jan 2012 23:13:42 +0800
Subject: [ANN] ngx_lua module v0.4.0 released!
In-Reply-To: <CAB4Tn6M9uDZ8j0jb24fiwj6VQ4GTaXgmeNHYXn=2=41TQ2N-Zw@mail.gmail.com>
References: <CAB4Tn6MGdGX1ZYXr99W1TWgkC6oZQLyq8Qv-XESXQzrZuUhM0g@mail.gmail.com>
	<CAB4Tn6M9uDZ8j0jb24fiwj6VQ4GTaXgmeNHYXn=2=41TQ2N-Zw@mail.gmail.com>
Message-ID: <CACPpB2yE66Pd8+3T+1ntb7Q805U1+iZ7BZev3388LKhFH4NJOA@mail.gmail.com>

great! have fun with ngx_lua~
--
blog: http://chenxiaoyu.org



2012/1/11 agentzh <agentzh at gmail.com>

> Hi, folks!
>
> After four months' active development of our contributors, I'm happy
> to announce the v0.4.0 release of our ngx_lua module. You can get the
> release tarball from the download page:
>
>    https://github.com/chaoslawful/lua-nginx-module/tags
>
> There's many many new features and bug fixes that you can find out in
> the following complete change log for this release (it is just too
> long to put here :P):
>
>    http://wiki.nginx.org/HttpLuaModule#v0.4.0
>
> Special thanks go to all of our contributors and users!
>
> This module embeds the Lua 5.1 interpreter or LuaJIT 2.0 into the
> nginx core and integrates the powerful Lua threads (aka Lua
> coroutines) into the nginx event model by means of nginx subrequests.
>
> Unlike Apache's mod_lua and Lighttpd's mod_magnet, Lua code written
> atop this module can be 100% non-blocking on network traffic as long
> as you use the ngx.location.capture or ngx.location.capture_multi
> interfaces to let the Nginx core do all your requests to mysql,
> postgresql, memcached, redis, upstream http web services, and etc etc
> etc.
>
> This module is also included and enabled by default in our
> ngx_openresty bundle: http://openresty.org/
>
> You can find the complete documentation for this module on the
> following wiki page:
>
>    http://wiki.nginx.org/HttpLuaModule
>
> And you always get the latest source code from the git repository here:
>
>    https://github.com/chaoslawful/lua-nginx-module
>
> Have fun!
> -agentzh
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120111/44fcedba/attachment.html>

From martinloy.uy at gmail.com  Wed Jan 11 15:31:35 2012
From: martinloy.uy at gmail.com (Martin Loy)
Date: Wed, 11 Jan 2012 13:31:35 -0200
Subject: mod_aclr2
In-Reply-To: <87sjjmxs2o.wl%appa@perusio.net>
References: <ED8F5676-8A10-4F60-94B8-C09812A4A954@nginx.com>
	<87sjjmxs2o.wl%appa@perusio.net>
Message-ID: <CAHbGeW6rZpAX0dMXp6ovHbUnXxWx5_eUBbVHMbtfbpMAPR_oSw@mail.gmail.com>

Would you explain the main difference between using your module and using
nginx as a reverse proxy and caching static content ?

Regards

M

On Wed, Jan 11, 2012 at 11:54 AM, Ant?nio P. P. Almeida <appa at perusio.net>wrote:

> On 11 Jan 2012 13h17 WET, defan at nginx.com wrote:
>
> > Hello!
> >
> > Ever wondered how to set up NGINX as a local proxy for Apache 2.x
> > with minimum efforts?
>
> Nice work Andrey. Opens up Nginx as a possibility for shared hosting
> (keep using .htaccess) and voids any resistance people might have trying
> Nginx out. It's sooo easy now :)
>
> Thanks,
> --- appa
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>



-- 
*Nunca hubo un amigo que hiciese un favor a un enano, ni un enemigo que le
hiciese un mal, que no se viese recompensado por entero.*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120111/4b76f0a3/attachment.html>

From defan at nginx.com  Wed Jan 11 18:09:38 2012
From: defan at nginx.com (Andrey Belov)
Date: Wed, 11 Jan 2012 22:09:38 +0400
Subject: mod_aclr2
In-Reply-To: <CAHbGeW6rZpAX0dMXp6ovHbUnXxWx5_eUBbVHMbtfbpMAPR_oSw@mail.gmail.com>
References: <ED8F5676-8A10-4F60-94B8-C09812A4A954@nginx.com>
	<87sjjmxs2o.wl%appa@perusio.net>
	<CAHbGeW6rZpAX0dMXp6ovHbUnXxWx5_eUBbVHMbtfbpMAPR_oSw@mail.gmail.com>
Message-ID: <10B30FF4-6B7B-4032-BBAF-495AF0C0E8EE@nginx.com>

Martin,

On Jan 11, 2012, at 7:31 PM, Martin Loy wrote:

> Would you explain the main difference between using your module and using nginx as a reverse proxy and caching static content ?

In some situations, mod_aclr2 helps to avoid excessive disk I/O,
especially for big static responses.

When nginx is used as a local reverse proxy and proxy_buffering is
enabled, every upstream reponse which can not fit into proxy buffers
may be temporarily saved on disk:

http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_buffering
http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_temp_path

For example, we have to process the request for 700mb video file,
without mod_aclr2:

1) nginx sends the request to apache
2) apache sends 700mb file back to nginx
3) nginx saves received file on disk (proxy_temp_path) and
starts sending it back to the client

and with mod_aclr2:

1) nginx sends the request to apache
2) apache sends a few bytes header with X-Accel-Redirect back to nginx
3) nginx starts sending the file directly from disk

You save your time and your disk at step #2.

Of course, you also may use proxy_cache along with mod_aclr2,
but if your disks are fast enough, probably you won't need
to do any caching at all.

Hope this helps.


> 
> Regards
> 
> M
> 
> On Wed, Jan 11, 2012 at 11:54 AM, Ant?nio P. P. Almeida <appa at perusio.net> wrote:
> On 11 Jan 2012 13h17 WET, defan at nginx.com wrote:
> 
> > Hello!
> >
> > Ever wondered how to set up NGINX as a local proxy for Apache 2.x
> > with minimum efforts?
> 
> Nice work Andrey. Opens up Nginx as a possibility for shared hosting
> (keep using .htaccess) and voids any resistance people might have trying
> Nginx out. It's sooo easy now :)
> 
> Thanks,
> --- appa
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
> 
> 
> 
> -- 
> Nunca hubo un amigo que hiciese un favor a un enano, ni un enemigo que le hiciese un mal, que no se viese recompensado por entero.
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120111/c8294306/attachment-0001.html>

From nginx-forum at nginx.us  Wed Jan 11 21:31:45 2012
From: nginx-forum at nginx.us (locojohn)
Date: Wed, 11 Jan 2012 16:31:45 -0500
Subject: Any way to deal with clients breaking the limit zone
In-Reply-To: <20110602093627.GZ42265@mdounin.ru>
References: <20110602093627.GZ42265@mdounin.ru>
Message-ID: <dc5b894dde405001691f7477923fa54a.NginxMailingListEnglish@forum.nginx.org>

Hello Maxim,

Sorry about highjacking this thread, but I think my question is
relevant.  In the nginx configuration I have:

        # limit simultaneous connections
        limit_conn_zone $binary_remote_addr zone=addr:1m;
        limit_conn addr 16;

But the error_log file says:

2012/01/11 21:41:39 [warn] 10140#0: *2959 limiting connections by zone
"addr", client: 127.0.0.1, server: , request: "GET / HTTP/1.0", host:
"mysite.com", referrer: "http://code.google.com/p/slowhttptest/"
2012/01/11 21:41:40 [warn] 10140#0: *2962 limiting connections by zone
"addr", client: 127.0.0.1, server: , request: "GET / HTTP/1.0", host:
"mysite.com", referrer: "http://code.google.com/p/slowhttptest/"

Question is, why 127.0.0.1 for the remote client address?  It should
have nothing to do with loopback host address.   It must be external
IP.

Any workaround?   Also, I noticed that "server:" is empty.   Any other
reason for that?   I have a dozen of virtual hosts configured with
fastcgi_pass to php-fpm.

Many thanks for your assistance.

nginx/1.1.12

Andrejs

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,203013,221055#msg-221055


From nginx-forum at nginx.us  Thu Jan 12 09:06:41 2012
From: nginx-forum at nginx.us (johnny)
Date: Thu, 12 Jan 2012 04:06:41 -0500
Subject: How to respond correctly to a POST request for a static file?
Message-ID: <8dc1162dd54944de31b5baec4d367743.NginxMailingListEnglish@forum.nginx.org>

hi,

I have a Facebook app hosted on nginx. Faceboook fetches the data by
issuing a POST request for index.html before embedding it in their
canvas. My server replies with 404, instead of the expected 405. If it
was 405, I could fix things by adding "error_page 405 =200 $uri;" to the
configuration file, but for some reason it's 404. 

I can think of two ways of correcting the situation:
1. Find a way to have the server respond with 405 and then add the above
line. This is probably the correct way.
2. Respond with 200 when a POST request is issued for index.html.

How do I implement either one of these?

thanks,
Johnny.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221059,221059#msg-221059


From albert at nn.iij4u.or.jp  Thu Jan 12 10:15:26 2012
From: albert at nn.iij4u.or.jp (shin fukuda)
Date: Thu, 12 Jan 2012 19:15:26 +0900
Subject: http requests consolidation
Message-ID: <20120112191526.607f450a02c1d8ac828c684b@nn.iij4u.or.jp>

Hi

We use Nginx as http proxy cache. 
Is Nginx supporting http requests consolidation ?
Or Is there a solution for that?



-- 
shin fukuda <albert at nn.iij4u.or.jp>


From nginx-forum at nginx.us  Thu Jan 12 13:08:04 2012
From: nginx-forum at nginx.us (abdelouahabb)
Date: Thu, 12 Jan 2012 08:08:04 -0500
Subject: nginx for windows 64 (not the cygwin)
Message-ID: <dfa4c8912a3627d4b9526974c1f7c09b.NginxMailingListEnglish@forum.nginx.org>

hi
i dont find the windows 64 bits build on the download section, what i've
found is only the cygwin version

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221072,221072#msg-221072


From nginx-forum at nginx.us  Thu Jan 12 13:19:53 2012
From: nginx-forum at nginx.us (abdelouahabb)
Date: Thu, 12 Jan 2012 08:19:53 -0500
Subject: nginx for windows 64 (not the cygwin)
In-Reply-To: <dfa4c8912a3627d4b9526974c1f7c09b.NginxMailingListEnglish@forum.nginx.org>
References: <dfa4c8912a3627d4b9526974c1f7c09b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <9ca413576f9133d92fe2e722079d5d88.NginxMailingListEnglish@forum.nginx.org>

abdelouahabb Wrote:
-------------------------------------------------------
> hi
> i dont find the windows 64 bits build on the
> download section, what i've found is only the
> cygwin version

*Edit: i speak about that because i've a Dell xps 15 to lean nginx with,
but sadly this is as i found, the most hated laptop from linux! all the
distro i've tested on it hangs, so i just want to stay on my seven 64
bits shipped with, and learn nginx without breaking my head about the OS
and its problems!
and the distribution (if it exists on win 64 bits) does it have the ssl
module activated?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221072,221073#msg-221073


From mdounin at mdounin.ru  Thu Jan 12 13:21:30 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 12 Jan 2012 17:21:30 +0400
Subject: http requests consolidation
In-Reply-To: <20120112191526.607f450a02c1d8ac828c684b@nn.iij4u.or.jp>
References: <20120112191526.607f450a02c1d8ac828c684b@nn.iij4u.or.jp>
Message-ID: <20120112132130.GD67687@mdounin.ru>

Hello!

On Thu, Jan 12, 2012 at 07:15:26PM +0900, shin fukuda wrote:

> Hi
> 
> We use Nginx as http proxy cache. 
> Is Nginx supporting http requests consolidation ?
> Or Is there a solution for that?

http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_cache_lock

Maxim Dounin


From mdounin at mdounin.ru  Thu Jan 12 13:26:21 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 12 Jan 2012 17:26:21 +0400
Subject: nginx for windows 64 (not the cygwin)
In-Reply-To: <9ca413576f9133d92fe2e722079d5d88.NginxMailingListEnglish@forum.nginx.org>
References: <dfa4c8912a3627d4b9526974c1f7c09b.NginxMailingListEnglish@forum.nginx.org>
	<9ca413576f9133d92fe2e722079d5d88.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120112132621.GE67687@mdounin.ru>

Hello!

On Thu, Jan 12, 2012 at 08:19:53AM -0500, abdelouahabb wrote:

> abdelouahabb Wrote:
> -------------------------------------------------------
> > hi
> > i dont find the windows 64 bits build on the
> > download section, what i've found is only the
> > cygwin version
> 
> *Edit: i speak about that because i've a Dell xps 15 to lean nginx with,
> but sadly this is as i found, the most hated laptop from linux! all the
> distro i've tested on it hangs, so i just want to stay on my seven 64
> bits shipped with, and learn nginx without breaking my head about the OS
> and its problems!
> and the distribution (if it exists on win 64 bits) does it have the ssl
> module activated?

There is no separate build available for 64-bit windows, use 
32-bit one, it should work (with usual limitations, see 
http://nginx.org/en/docs/windows.html).

Alternatively, you may build it yourself as specified here:

http://nginx.org/en/docs/howto_build_on_win32.html

Or use your favorite virtual machine to run nginx under your 
favorite *nix variant.

Maxim Dounin


From alabdelouahab at gmail.com  Thu Jan 12 13:30:23 2012
From: alabdelouahab at gmail.com (aliane abdelouahab)
Date: Thu, 12 Jan 2012 14:30:23 +0100
Subject: nginx for windows 64 (not the cygwin)
In-Reply-To: <20120112132621.GE67687@mdounin.ru>
References: <dfa4c8912a3627d4b9526974c1f7c09b.NginxMailingListEnglish@forum.nginx.org>
	<9ca413576f9133d92fe2e722079d5d88.NginxMailingListEnglish@forum.nginx.org>
	<20120112132621.GE67687@mdounin.ru>
Message-ID: <CAP4kiWpo2iZiwzGqJW9g6fp0Xv4vuv3_bwYR9tAUa_QG8UyCMw@mail.gmail.com>

2012/1/12 Maxim Dounin <mdounin at mdounin.ru>

> Hello!
>
> On Thu, Jan 12, 2012 at 08:19:53AM -0500, abdelouahabb wrote:
>
> > abdelouahabb Wrote:
> > -------------------------------------------------------
> > > hi
> > > i dont find the windows 64 bits build on the
> > > download section, what i've found is only the
> > > cygwin version
> >
> > *Edit: i speak about that because i've a Dell xps 15 to lean nginx with,
> > but sadly this is as i found, the most hated laptop from linux! all the
> > distro i've tested on it hangs, so i just want to stay on my seven 64
> > bits shipped with, and learn nginx without breaking my head about the OS
> > and its problems!
> > and the distribution (if it exists on win 64 bits) does it have the ssl
> > module activated?
>
> There is no separate build available for 64-bit windows, use
> 32-bit one, it should work (with usual limitations, see
> http://nginx.org/en/docs/windows.html).
>
> Alternatively, you may build it yourself as specified here:
>
> http://nginx.org/en/docs/howto_build_on_win32.html
>
> Or use your favorite virtual machine to run nginx under your
> favorite *nix variant.
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>

thank you :)
sorry, but i've not Visual studio, am running codes under python, it's free
:p
is there someone that can provide us a 64 bits version???
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/d8d2e60b/attachment.html>

From ne at vbart.ru  Thu Jan 12 13:35:59 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Thu, 12 Jan 2012 17:35:59 +0400
Subject: nginx for windows 64 (not the cygwin)
In-Reply-To: <CAP4kiWpo2iZiwzGqJW9g6fp0Xv4vuv3_bwYR9tAUa_QG8UyCMw@mail.gmail.com>
References: <dfa4c8912a3627d4b9526974c1f7c09b.NginxMailingListEnglish@forum.nginx.org>
	<20120112132621.GE67687@mdounin.ru>
	<CAP4kiWpo2iZiwzGqJW9g6fp0Xv4vuv3_bwYR9tAUa_QG8UyCMw@mail.gmail.com>
Message-ID: <201201121735.59770.ne@vbart.ru>

On Thursday 12 January 2012 17:30:23 aliane abdelouahab wrote:
[...]
> 
> thank you :)
> sorry, but i've not Visual studio, am running codes under python, it's free
> 
> :p
> 
> is there someone that can provide us a 64 bits version???

Sorry, but what's wrong with 32'bit version?

There's no difference for your purposes.


wbr, Valentin V. Bartenev


From alabdelouahab at gmail.com  Thu Jan 12 13:40:20 2012
From: alabdelouahab at gmail.com (aliane abdelouahab)
Date: Thu, 12 Jan 2012 14:40:20 +0100
Subject: nginx for windows 64 (not the cygwin)
In-Reply-To: <201201121735.59770.ne@vbart.ru>
References: <dfa4c8912a3627d4b9526974c1f7c09b.NginxMailingListEnglish@forum.nginx.org>
	<20120112132621.GE67687@mdounin.ru>
	<CAP4kiWpo2iZiwzGqJW9g6fp0Xv4vuv3_bwYR9tAUa_QG8UyCMw@mail.gmail.com>
	<201201121735.59770.ne@vbart.ru>
Message-ID: <CAP4kiWoT8R98GL7qodt1eE0RmU0bk9W_MSoDbZVotg9KqMFJWw@mail.gmail.com>

2012/1/12 Valentin V. Bartenev <ne at vbart.ru>

> On Thursday 12 January 2012 17:30:23 aliane abdelouahab wrote:
> [...]
> >
> > thank you :)
> > sorry, but i've not Visual studio, am running codes under python, it's
> free
> >
> > :p
> >
> > is there someone that can provide us a 64 bits version???
>
> Sorry, but what's wrong with 32'bit version?
>
> There's no difference for your purposes.
>
>
> wbr, Valentin V. Bartenev
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>

i run under 8 giga ram, so i just wanted to test the limits of a server
running under 64 bits, my thesis is to make a webserver using tornado (also
not officialy supported on windows) and as load balancer and reverse proxy
i'll use nginx (as suggested in tornado documentation)
and what about the 32 bits build, does it contain the basic modules +
extras?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/895e11b0/attachment.html>

From jimmysongo at gmail.com  Thu Jan 12 13:52:15 2012
From: jimmysongo at gmail.com (Jimmysongo)
Date: Thu, 12 Jan 2012 21:52:15 +0800
Subject: mod_aclr2
In-Reply-To: <10B30FF4-6B7B-4032-BBAF-495AF0C0E8EE@nginx.com>
References: <ED8F5676-8A10-4F60-94B8-C09812A4A954@nginx.com>
	<87sjjmxs2o.wl%appa@perusio.net>
	<CAHbGeW6rZpAX0dMXp6ovHbUnXxWx5_eUBbVHMbtfbpMAPR_oSw@mail.gmail.com>
	<10B30FF4-6B7B-4032-BBAF-495AF0C0E8EE@nginx.com>
Message-ID: <AB6D54E3-C023-4009-A25E-F8B87C2225EA@gmail.com>

use ramdisk?tmpfs



? 2012-1-12?2:09?Andrey Belov <defan at nginx.com> ???

> Martin,
> 
> On Jan 11, 2012, at 7:31 PM, Martin Loy wrote:
> 
>> Would you explain the main difference between using your module and using nginx as a reverse proxy and caching static content ?
> 
> In some situations, mod_aclr2 helps to avoid excessive disk I/O,
> especially for big static responses.
> 
> When nginx is used as a local reverse proxy and proxy_buffering is
> enabled, every upstream reponse which can not fit into proxy buffers
> may be temporarily saved on disk:
> 
> http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_buffering
> http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_temp_path
> 
> For example, we have to process the request for 700mb video file,
> without mod_aclr2:
> 
> 1) nginx sends the request to apache
> 2) apache sends 700mb file back to nginx
> 3) nginx saves received file on disk (proxy_temp_path) and
> starts sending it back to the client
> 
> and with mod_aclr2:
> 
> 1) nginx sends the request to apache
> 2) apache sends a few bytes header with X-Accel-Redirect back to nginx
> 3) nginx starts sending the file directly from disk
> 
> You save your time and your disk at step #2.
> 
> Of course, you also may use proxy_cache along with mod_aclr2,
> but if your disks are fast enough, probably you won't need
> to do any caching at all.
> 
> Hope this helps.
> 
> 
>> 
>> Regards
>> 
>> M
>> 
>> On Wed, Jan 11, 2012 at 11:54 AM, Ant?nio P. P. Almeida <appa at perusio.net> wrote:
>> On 11 Jan 2012 13h17 WET, defan at nginx.com wrote:
>> 
>> > Hello!
>> >
>> > Ever wondered how to set up NGINX as a local proxy for Apache 2.x
>> > with minimum efforts?
>> 
>> Nice work Andrey. Opens up Nginx as a possibility for shared hosting
>> (keep using .htaccess) and voids any resistance people might have trying
>> Nginx out. It's sooo easy now :)
>> 
>> Thanks,
>> --- appa
>> 
>> _______________________________________________
>> nginx mailing list
>> nginx at nginx.org
>> http://mailman.nginx.org/mailman/listinfo/nginx
>> 
>> 
>> 
>> -- 
>> Nunca hubo un amigo que hiciese un favor a un enano, ni un enemigo que le hiciese un mal, que no se viese recompensado por entero.
>> _______________________________________________
>> nginx mailing list
>> nginx at nginx.org
>> http://mailman.nginx.org/mailman/listinfo/nginx
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/7de087ec/attachment-0001.html>

From ne at vbart.ru  Thu Jan 12 14:03:21 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Thu, 12 Jan 2012 18:03:21 +0400
Subject: nginx for windows 64 (not the cygwin)
In-Reply-To: <CAP4kiWoT8R98GL7qodt1eE0RmU0bk9W_MSoDbZVotg9KqMFJWw@mail.gmail.com>
References: <dfa4c8912a3627d4b9526974c1f7c09b.NginxMailingListEnglish@forum.nginx.org>
	<201201121735.59770.ne@vbart.ru>
	<CAP4kiWoT8R98GL7qodt1eE0RmU0bk9W_MSoDbZVotg9KqMFJWw@mail.gmail.com>
Message-ID: <201201121803.21728.ne@vbart.ru>

On Thursday 12 January 2012 17:40:20 aliane abdelouahab wrote:
> i run under 8 giga ram, so i just wanted to test the limits of a server
> running under 64 bits, my thesis is to make a webserver using tornado (also
> not officialy supported on windows) and as load balancer and reverse proxy
> i'll use nginx (as suggested in tornado documentation)

http://nginx.org/en/docs/windows.html - this limitations relate to all windows 
builds, not only 32-bit. Don't expect high performance or more than 1024 
simultaneous connections.

> and what about the 32 bits build, does it contain the basic modules +
> extras?

What "extras" do you mean exactly? Some of nginx modules don't work properly on 
windows 7, and, probably, most of 3'rd party modules even won't compile.

wbr, Valentin V. Bartenev


From zhuzhaoyuan at gmail.com  Thu Jan 12 15:35:06 2012
From: zhuzhaoyuan at gmail.com (Joshua Zhu)
Date: Thu, 12 Jan 2012 23:35:06 +0800
Subject: [ANNOUNCE] Tengine-1.2.2
Message-ID: <CAOKiq2C8czdC8+XBOoxuB04cq+g-jVooz2owt=5WLPDnx2x1OQ@mail.gmail.com>

Hi,

We are pleased to announce that Tengine-1.2.2 has been released. And you
can download it here: http://tengine.taobao.org/download/tengine.tar.gz
(For those who don't know Tengine, it is a distribution of Nginx, including
quite a few useful patches and modules developed by a group software
engineers from Taobao)

In this release, we introduced the 'input body filter' mechanism. So it's
very handy to a developer who wants to write a Web Application Firewall. We
merged the changes from the latest stable Nginx, i.e. Tengine is based on
Nginx-1.0.11 now. There are also some bug fixes in this release. More
detailed changelog is available at http://tengine.taobao.org/changelog.html

Have fun, folks!

Regards,

-- 
Joshua Zhu
Senior Software Engineer
Server Platforms Team at Taobao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/d2a8b5e8/attachment.html>

From zhuzhaoyuan at gmail.com  Thu Jan 12 15:36:38 2012
From: zhuzhaoyuan at gmail.com (Joshua Zhu)
Date: Thu, 12 Jan 2012 23:36:38 +0800
Subject: [PATCH 0 of 2] CPU affinity fixes
Message-ID: <CAOKiq2C5y8nsLXdpo-CyQ4BGZvuqAtwepCAKkGTWwmF3Sx4pAQ@mail.gmail.com>

Hi,

Here are two patches for the CPU affinity setting feature.
Code review and testing are appreciated.

Regards,

-- 
Joshua Zhu
Senior Software Engineer
Server Platforms Team at Taobao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/b8320d7c/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bugfix.patch
Type: text/x-patch
Size: 6447 bytes
Desc: not available
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/b8320d7c/attachment.bin>

From zhuzhaoyuan at gmail.com  Thu Jan 12 15:37:00 2012
From: zhuzhaoyuan at gmail.com (Joshua Zhu)
Date: Thu, 12 Jan 2012 23:37:00 +0800
Subject: [PATCH 1 of 2] fix sched_setaffinity(2) call failure
Message-ID: <CAOKiq2D2hwCFy3+9wFeU8ZVxtHRgU3Nrj4-UafSuYhgn8AzCOw@mail.gmail.com>

Hi,

'cpu_set_t' should be used instead of 'long' when calling
sched_setaffinity(2). Otherwise it may fail on some Linux systems.
Please see the attachment for more detail.

Regards,

-- 
Joshua Zhu
Senior Software Engineer
Server Platforms Team at Taobao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/b5c3cf6e/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bugfix.patch
Type: text/x-patch
Size: 6447 bytes
Desc: not available
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/b5c3cf6e/attachment-0001.bin>

From zhuzhaoyuan at gmail.com  Thu Jan 12 15:37:15 2012
From: zhuzhaoyuan at gmail.com (Joshua Zhu)
Date: Thu, 12 Jan 2012 23:37:15 +0800
Subject: [PATCH 2 of 2] add support for setting worker processes and CPU
	affinity automatically
Message-ID: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>

Hi,

It's a little bit inconvenient to set the CPU affinities with
many-core-system. For example, to fully utilize a 16-core Linux box, we
have to write something like this:
worker_processes 16;
worker_cpu_affinity  1000000000000000 0100000000000000 0010000000000000
0001000000000000 0000100000000000 0000010000000000 0000001000000000
0000000100000000 0000000010000000 0000000001000000 0000000000100000
0000000000010000 0000000000001000 0000000000000100 0000000000000010
0000000000000001;

This patch will make life much easier:
worker_processes auto;
worker_cpu_affinity auto;

Please see the attachment for more detail.

Regards,

-- 
Joshua Zhu
Senior Software Engineer
Server Platforms Team at Taobao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/692776cf/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: auto.patch
Type: text/x-patch
Size: 5161 bytes
Desc: not available
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/692776cf/attachment.bin>

From maxim at nginx.com  Thu Jan 12 16:07:03 2012
From: maxim at nginx.com (Maxim Konovalov)
Date: Thu, 12 Jan 2012 20:07:03 +0400
Subject: [PATCH 2 of 2] add support for setting worker processes and CPU
	affinity automatically
In-Reply-To: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>
References: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>
Message-ID: <4F0F0527.3030104@nginx.com>

Hi Joshua,

nice work indeed, thanks!

Do you have a practical example when such explicit binding
does better work than OS scheduler?

-- 
Maxim Konovalov
+7 (910) 4293178
http://nginx.com/


From mdounin at mdounin.ru  Thu Jan 12 16:11:29 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 12 Jan 2012 20:11:29 +0400
Subject: Any way to deal with clients breaking the limit zone
In-Reply-To: <dc5b894dde405001691f7477923fa54a.NginxMailingListEnglish@forum.nginx.org>
References: <20110602093627.GZ42265@mdounin.ru>
	<dc5b894dde405001691f7477923fa54a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120112161129.GI67687@mdounin.ru>

Hello!

On Wed, Jan 11, 2012 at 04:31:45PM -0500, locojohn wrote:

> Hello Maxim,
> 
> Sorry about highjacking this thread, but I think my question is
> relevant.  In the nginx configuration I have:
> 
>         # limit simultaneous connections
>         limit_conn_zone $binary_remote_addr zone=addr:1m;
>         limit_conn addr 16;
> 
> But the error_log file says:
> 
> 2012/01/11 21:41:39 [warn] 10140#0: *2959 limiting connections by zone
> "addr", client: 127.0.0.1, server: , request: "GET / HTTP/1.0", host:
> "mysite.com", referrer: "http://code.google.com/p/slowhttptest/"
> 2012/01/11 21:41:40 [warn] 10140#0: *2962 limiting connections by zone
> "addr", client: 127.0.0.1, server: , request: "GET / HTTP/1.0", host:
> "mysite.com", referrer: "http://code.google.com/p/slowhttptest/"
> 
> Question is, why 127.0.0.1 for the remote client address?  It should
> have nothing to do with loopback host address.   It must be external
> IP.

Obvious reason is: because connection was got from 127.0.0.1.

> Any workaround?   Also, I noticed that "server:" is empty.   Any other
> reason for that?   I have a dozen of virtual hosts configured with
> fastcgi_pass to php-fpm.

Obvious reason is: the request was matched by a server{} block 
without server_name set.  See here:

http://nginx.org/en/docs/http/request_processing.html

for more details on how nginx matches requests to servers.

Maxim Dounin


From nginx-forum at nginx.us  Thu Jan 12 16:21:40 2012
From: nginx-forum at nginx.us (voidandany)
Date: Thu, 12 Jan 2012 11:21:40 -0500
Subject: Protect a specific php file
Message-ID: <d14ef89255215a46b37abc5ec00190d5.NginxMailingListEnglish@forum.nginx.org>

Hello, 

On my server I have a php app in a subfolder of my root, so basically
I've made the following nginx configuration : 


server {
        listen       80;
        server_name  www.domain.fr;
        root         /var/www/domain.fr;
        access_log      /var/log/nginx/access.log ;
        location /myapp {
                location ~ \.php$ {
                        include         /etc/nginx/fastcgi_params;
                        fastcgi_param   SCRIPT_FILENAME 
$document_root$fastcgi_script_name;
                        fastcgi_pass    127.0.0.1:9000;
                }
        }
}

And it's work now in this app I want to protect the index.php file and
only this one, so I try this : 


        location /myapp {
                location ^~ /myapp/index.php {
                        auth_basic            "Private Section";
                        auth_basic_user_file  $document_root/.htpasswd;
                        location ~ \.php$ {
                                include        
/etc/nginx/fastcgi_params;
                                fastcgi_param   SCRIPT_FILENAME 
$document_root$fastcgi_script_name;
                                fastcgi_pass    127.0.0.1:9000;
                        }
                }
                location ~ \.php$ {
                        include         /etc/nginx/fastcgi_params;
                        fastcgi_param   SCRIPT_FILENAME 
$document_root$fastcgi_script_name;
                        fastcgi_pass    127.0.0.1:9000;
                }
        }

I had to duplicate the php location, is there an another way ?

It work with IE8 but not with Chrome !!

With Chrome the complete URL (http://www.domain.fr/myapp/index.php) is
correctly protected but not this one http://www.domain.fr/myapp/
In this case the php file is downloaded...

Any idea why ? Any solution ?

Thanks a lot

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221094,221094#msg-221094


From zhuzhaoyuan at gmail.com  Thu Jan 12 16:41:58 2012
From: zhuzhaoyuan at gmail.com (Joshua Zhu)
Date: Fri, 13 Jan 2012 00:41:58 +0800
Subject: [PATCH 2 of 2] add support for setting worker processes and CPU
	affinity automatically
In-Reply-To: <4F0F0527.3030104@nginx.com>
References: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>
	<4F0F0527.3030104@nginx.com>
Message-ID: <CAOKiq2DR+kjRd6uUSwjOJdaeutRmivCyYxndLkxaWu03cS7Nig@mail.gmail.com>

Hi Maxim,

On Fri, Jan 13, 2012 at 12:07 AM, Maxim Konovalov <maxim at nginx.com> wrote:

> Hi Joshua,
>
> nice work indeed, thanks!
>
> Do you have a practical example when such explicit binding
> does better work than OS scheduler?
>
> --
> Maxim Konovalov
> +7 (910) 4293178
> http://nginx.com/
>

Thanks!
Actually we did a test against the CPU affinity feature before and found it
was a little bit better than what the OS scheduler did, not too much
though. This is as expected since there will be less CPU cache missing with
this feature turned on.

Regards,

-- 
Joshua Zhu
Senior Software Engineer
Server Platforms Team at Taobao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/9c7caadb/attachment.html>

From gwei1016 at gmail.com  Thu Jan 12 08:41:32 2012
From: gwei1016 at gmail.com (=?gb2312?B?uN/OoQ==?=)
Date: Thu, 12 Jan 2012 16:41:32 +0800
Subject: =?UTF-8?Q?=E3=80=90help_to_configure_nginx_proxy=E3=80=91?=
Message-ID: <089b01ccd106$010efc30$032cf490$@gmail.com>

         Hello:

                            I want to config a proxy server to access https
backend, but nginx return 400, I don?t how to config

 

 

 

         My config is:

 

#user  nobody;

worker_processes  3;

 

error_log  logs/error.log;

pid        logs/nginx.pid;

 

 

events {

    worker_connections  1024;

}

 

http {

    include       mime.types;

    default_type  application/octet-stream;

    log_format test '$http_clientip $http_baeuser $remote_user [$time_local]
'

                       '$status $request_length $bytes_sent $request_time '

                       '0 0 0 - - $request - -';

    access_log  logs/access.log  test;

 

    sendfile        on;

    keepalive_timeout  5;

    proxy_connect_timeout    5;

    proxy_read_timeout       60;

    proxy_send_timeout       60;

    proxy_buffer_size        16k;

    proxy_buffers            4 64k;

    proxy_busy_buffers_size  128k;

    proxy_temp_file_write_size 128k;

 

 

    #server {

     #   listen       8086;

        #       resolver 10.65.0.231;

        #location / {

         #              proxy_pass http://$http_host$request_uri;

          #             proxy_next_upstream http_502 http_504 error timeout
invalid_header;

           #            proxy_set_header Host  $host;

            #           proxy_set_header X-Forwarded-For  $remote_addr;

             #          proxy_set_header cookie $http_cookie;

        #}

    #}

    # HTTPS server

    map $scheme $msiis      { http off; https on; }    

    server {

        listen      443; 

        resolver 10.65.0.231;

        ssl on;

        ssl_certificate      /home/work/imgw/nginx/ssl/cert.crt;

        ssl_certificate_key  /home/work/imgw/nginx/ssl/cert.key;

        ssl_verify_client off;

        #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 / {

                proxy_pass $scheme://$host$request_uri;

                proxy_set_header X-Forwarded-Proto https;

                proxy_set_header Host  $http_host;

                proxy_set_header X-Forwarded-By   $server_addr:$server_port;

                proxy_set_header X-Forwarded-For  $remote_addr;

                proxy_set_header        Front-End-Https   $msiis;

                proxy_set_header cookie $http_cookie;

 

 

        }

    }

}

 

 

         Thanks to your help!

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/342a8023/attachment-0001.html>

From appa at perusio.net  Thu Jan 12 16:52:34 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 12 Jan 2012 16:52:34 +0000
Subject: Protect a specific php file
In-Reply-To: <d14ef89255215a46b37abc5ec00190d5.NginxMailingListEnglish@forum.nginx.org>
References: <d14ef89255215a46b37abc5ec00190d5.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87ehv4yib1.wl%appa@perusio.net>

On 12 Jan 2012 16h21 WET, nginx-forum at nginx.us wrote:

> Hello, 
>
> Any idea why ? Any solution ?

Yes. Duplicate the FCGI setup in the /myapp/index.php location.

                location ^~ /myapp/index.php {
                    auth_basic "Private Section";
                    auth_basic_user_file  $document_root/.htpasswd;
                    include /etc/nginx/fastcgi_params;
                    fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
                    fastcgi_pass    127.0.0.1:9000;
                }

--- appa



From ben.lancaster at holler.co.uk  Thu Jan 12 15:15:18 2012
From: ben.lancaster at holler.co.uk (Ben Lancaster)
Date: Thu, 12 Jan 2012 15:15:18 +0000
Subject: Programmatic access to Nginx fcgi/proxy cache?
Message-ID: <OFD9E71ECE.515D64F4-ON80257983.00537F94-80257983.0053CC64@lionresources.com>

An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/d981f876/attachment.html>

From appa at perusio.net  Thu Jan 12 17:27:45 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 12 Jan 2012 17:27:45 +0000
Subject: Programmatic access to Nginx fcgi/proxy cache?
In-Reply-To: <OFD9E71ECE.515D64F4-ON80257983.00537F94-80257983.0053CC64@lionresources.com>
References: <OFD9E71ECE.515D64F4-ON80257983.00537F94-80257983.0053CC64@lionresources.com>
Message-ID: <87d3aoygoe.wl%appa@perusio.net>

On 12 Jan 2012 15h15 WET, ben.lancaster at holler.co.uk wrote:

You can process the cache files. They're regular files.

Here's a script to inspect the cache:

https://github.com/perusio/nginx-cache-inspector
 
another for purging:

https://github.com/perusio/nginx-cache-purge

There's also a 3rd party module for cache purging. AFAIK doesn't
support wildcards:

https://github.com/FRiCKLE/ngx_cache_purge

Just make a HTTP request from your app.

--- appa


From ne at vbart.ru  Thu Jan 12 17:53:56 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Thu, 12 Jan 2012 21:53:56 +0400
Subject: [PATCH 2 of 2] add support for setting worker processes and CPU
	affinity automatically
In-Reply-To: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>
References: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>
Message-ID: <201201122153.56491.ne@vbart.ru>

On Thursday 12 January 2012 19:37:15 Joshua Zhu wrote:
> It's a little bit inconvenient to set the CPU affinities with
> many-core-system. For example, to fully utilize a 16-core Linux box, we
> have to write something like this:
> worker_processes 16;
> worker_cpu_affinity  1000000000000000 0100000000000000 0010000000000000
> 0001000000000000 0000100000000000 0000010000000000 0000001000000000
> 0000000100000000 0000000010000000 0000000001000000 0000000000100000
> 0000000000010000 0000000000001000 0000000000000100 0000000000000010
> 0000000000000001;
> 
> This patch will make life much easier:
> worker_processes auto;
> worker_cpu_affinity auto;
> 
> Please see the attachment for more detail.
> 

You use "ngx_ncpu" global variable and for Linux you set it by 
sysconf(_SC_NPROCESSORS_ONLN).

On freeBSD and MacOS systems it set by sysctls(hw.ncpu), and it is equal
to 1 on others.

Which means that your "auto" settings are very inaccurate. Sometimes the
"worker_processes" directive will be equal to 2 * cores or more, sometimes
it will be equal to 1, and if you're lucky, then it will be equal to real
CPU cores.

So, I don't think that this patch makes life easier. In fact, it creates a
false sense of optimal worker_processes setting. Even if ngx_ncpu would be
always equal to real hardware cores number, it wouldn't be optimal for every
use cases.

wbr, Valentin V. Bartenev


From karljohn at gmail.com  Thu Jan 12 18:30:43 2012
From: karljohn at gmail.com (Karl Johnson)
Date: Thu, 12 Jan 2012 18:30:43 +0000
Subject: help with ssl config nginx -> thin -> rails
Message-ID: <CAEDYxyzNzYaOwz5De6rNr+hT1OtOtW4O94=Z917zvYkEiSruMQ@mail.gmail.com>

I'm lost in a sea of configurations and every change I make ends up causing
a different problem in relation to serving over HTTPS. I've tried all kinds
of different combos of the headers. I'm using nginx in front of thin web
servers for rails.

With the below config, serving static content works. Serving rails pages
now just bombs over HTTPS with a 400. The prior config I had caused an
endless redirect.

Being a relative newbie to nginx and rails, my guess is that it has
something to do with the thin servers serving over the 400x ports, but the
upstream below just referencing 443. How should this be configured to serve
dynamic rails content over SSL when you have 20 nodes on thin started?

Thanks for any help!

upstream developmentmode{
        server 127.0.0.1:4000;
        server 127.0.0.1:4001;
        server 127.0.0.1:4002;
        server 127.0.0.1:4003;
        server 127.0.0.1:4004;
        server 127.0.0.1:4005;
        server 127.0.0.1:4006;
        server 127.0.0.1:4007;
        server 127.0.0.1:4008;
        server 127.0.0.1:4009;
        server 127.0.0.1:4010;
        server 127.0.0.1:4011;
        server 127.0.0.1:4012;
        server 127.0.0.1:4013;
        server 127.0.0.1:4014;
        server 127.0.0.1:4015;
        server 127.0.0.1:4016;
        server 127.0.0.1:4017;
        server 127.0.0.1:4018;
        server 127.0.0.1:4019;
    }

upstream devmode-secure{
server 127.0.0.1:443;
}

server {
listen 80;
server_name govenga.com www.govenga.com;
server_name 50.56.121.244;
access_log /var/www/dev/log/access.log;
        error_log /var/www/dev/log/error.log;

        client_max_body_size 50M;
        client_body_buffer_size 512k;

        root   /var/www/dev/public/;
        index  index.html;
        location ~*
^.+.(jpg|jpeg|gif|png|ico|css|zip|tgz|gz|rar|bz2|doc|xls|exe|ppt|txt|tar|mid|midi|wav|bmp|rtf|ttf|woff|svg|eot)$
{
        root   /var/www/dev/public/;
                expires           10y;
                add_header Cache-Control public;
        }
location ~ ^(/javascripts/.*)$ {
        root   /var/www/dev/public/;
                expires           10y;
                add_header Cache-Control public;
        }
location / {
proxy_set_header  X-Real-IP  $remote_addr;
proxy_set_header  X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header Host $host:$proxy_port;
proxy_redirect off;

 if (-f $request_filename/index.html) {
        rewrite (.*) $1/index.html break;
      }
      if (-f $request_filename.html) {
        rewrite (.*) $1.html break;
      }
      if (!-f $request_filename) {
        proxy_pass http://developmentmode;
        break;
      }
}
}
server {
listen 443;
server_name govenga.com www.govenga.com;
ssl                  on;
ssl_certificate      /etc/nginx/certificates/new/govenga.com.crt;
        ssl_certificate_key  /etc/nginx/certificates/new/govenga_new.key;
ssl_client_certificate /etc/nginx/certificates/new/gd_bundle.crt;
        keepalive_timeout    70;
#
        access_log /var/www/dev/log/access.log;
        error_log /var/www/dev/log/error.log;
#
        client_max_body_size 50M;
        client_body_buffer_size 512k;
#
        root   /var/www/dev/public/;
        index  index.html;
        location ~*
^.+.(jpg|jpeg|gif|png|ico|css|zip|tgz|gz|rar|bz2|doc|xls|exe|ppt|txt|tar|mid|midi|wav|bmp|rtf|ttf|woff|svg|eot)$
{
        root   /var/www/dev/public/;
                expires           1y;
                add_header Cache-Control public;
        }
  location / {
      proxy_set_header X-Real-IP $remote_addr;
      proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
      proxy_set_header  Host $http_host;
      proxy_set_header X-Forwarded-Proto https;
      proxy_redirect off;
      proxy_set_header  X-Url-Scheme $scheme;

      if (-f $request_filename/index.html) {
        rewrite (.*) $1/index.html break;
      }
      if (-f $request_filename.html) {
        rewrite (.*) $1.html break;
      }
      if (!-f $request_filename) {
        proxy_pass https://devmode-secure;
        break;
      }
    }
}
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/75e8de79/attachment.html>

From greg at qualcomm.com  Thu Jan 12 18:39:51 2012
From: greg at qualcomm.com (Richardson, Greg)
Date: Thu, 12 Jan 2012 18:39:51 +0000
Subject: [ANNOUNCE] Tengine-1.2.2
In-Reply-To: <CAOKiq2C8czdC8+XBOoxuB04cq+g-jVooz2owt=5WLPDnx2x1OQ@mail.gmail.com>
References: <CAOKiq2C8czdC8+XBOoxuB04cq+g-jVooz2owt=5WLPDnx2x1OQ@mail.gmail.com>
Message-ID: <C45999C885E7CA46B0E5847C6D1F0CC8DE2F9E@nasanexd01b.na.qualcomm.com>

Good stuff!  The following feature caught my eye, but I didn't notice any documentation / samples (though I might have missed them).  Is this feature-set present in 1.2.2?

-Combines multiple CSS or JavasScript requests into one request to reduce the downloading time

Side note, the Server header on tengine.taobao.org is 'Apache'.  Say what!?!  :)

From: nginx-bounces at nginx.org [mailto:nginx-bounces at nginx.org] On Behalf Of Joshua Zhu
Sent: Thursday, January 12, 2012 7:35 AM
To: nginx at nginx.org
Subject: [ANNOUNCE] Tengine-1.2.2

Hi,

We are pleased to announce that Tengine-1.2.2 has been released. And you can download it here: http://tengine.taobao.org/download/tengine.tar.gz
(For those who don't know Tengine, it is a distribution of Nginx, including quite a few useful patches and modules developed by a group software engineers from Taobao)

In this release, we introduced the 'input body filter' mechanism. So it's very handy to a developer who wants to write a Web Application Firewall. We merged the changes from the latest stable Nginx, i.e. Tengine is based on Nginx-1.0.11 now. There are also some bug fixes in this release. More detailed changelog is available at http://tengine.taobao.org/changelog.html

Have fun, folks!

Regards,

--
Joshua Zhu
Senior Software Engineer
Server Platforms Team at Taobao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/9041806e/attachment-0001.html>

From cliff at develix.com  Thu Jan 12 18:42:46 2012
From: cliff at develix.com (Cliff Wells)
Date: Thu, 12 Jan 2012 10:42:46 -0800
Subject: How to respond correctly to a POST request for a static file?
In-Reply-To: <8dc1162dd54944de31b5baec4d367743.NginxMailingListEnglish@forum.nginx.org>
References: <8dc1162dd54944de31b5baec4d367743.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <1326393766.10162.19.camel@portable-evil>

http://invalidlogic.com/2011/04/12/serving-static-content-via-post-from-nginx/




On Thu, 2012-01-12 at 04:06 -0500, johnny wrote:
> hi,
> 
> I have a Facebook app hosted on nginx. Faceboook fetches the data by
> issuing a POST request for index.html before embedding it in their
> canvas. My server replies with 404, instead of the expected 405. If it
> was 405, I could fix things by adding "error_page 405 =200 $uri;" to the
> configuration file, but for some reason it's 404. 
> 
> I can think of two ways of correcting the situation:
> 1. Find a way to have the server respond with 405 and then add the above
> line. This is probably the correct way.
> 2. Respond with 200 when a POST request is issued for index.html.
> 
> How do I implement either one of these?
> 
> thanks,
> Johnny.
> 
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221059,221059#msg-221059
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx



From reallfqq-nginx at yahoo.fr  Thu Jan 12 18:43:26 2012
From: reallfqq-nginx at yahoo.fr (B.R.)
Date: Thu, 12 Jan 2012 13:43:26 -0500
Subject: Error log file
In-Reply-To: <CALqce=3VxJPE86qn6k9SZRLJryB6JJtkkPP96BWSA1Z8qLJLbg@mail.gmail.com>
References: <CALqce=3VxJPE86qn6k9SZRLJryB6JJtkkPP96BWSA1Z8qLJLbg@mail.gmail.com>
Message-ID: <CALqce=1h-B67wmufmsUm=RQiSKxtPOyrbE_oeiKtif9GXDsvoQ@mail.gmail.com>

I don't know if the message was delivered the first time.
---
*B. R.*


On Thu, Jan 5, 2012 at 01:33, B.R. <reallfqq-nginx at yahoo.fr> wrote:

> Hello,
>
> Using Nginx 1.0.11, I changed the error log file in my configuration to
> /var/log/nginx.error.log.
> The new log file gets filled as it is supposed to be, but the old file
> /nginx-repo/logs/error.log keeps getting recreated.
>
> I would add that I checked the configuration so there is no conflicting
> entry.
>
> Btw, why does nginx doesn't use the 'standard' repos /var/log for the logs
> and /var/run for the PID file by default? It adds to the entropy...
> ---
> *B. R.*
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120112/2ec16540/attachment.html>

From appa at perusio.net  Thu Jan 12 18:45:21 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 12 Jan 2012 18:45:21 +0000
Subject: [ANNOUNCE] Tengine-1.2.2
In-Reply-To: <C45999C885E7CA46B0E5847C6D1F0CC8DE2F9E@nasanexd01b.na.qualcomm.com>
References: <CAOKiq2C8czdC8+XBOoxuB04cq+g-jVooz2owt=5WLPDnx2x1OQ@mail.gmail.com>
	<C45999C885E7CA46B0E5847C6D1F0CC8DE2F9E@nasanexd01b.na.qualcomm.com>
Message-ID: <87aa5syd32.wl%appa@perusio.net>

On 12 Jan 2012 18h39 WET, greg at qualcomm.com wrote:

> Good stuff!  The following feature caught my eye, but I didn't
> notice any documentation / samples (though I might have missed
> them).  Is this feature-set present in 1.2.2?
>
> -Combines multiple CSS or JavasScript requests into one request to
> reduce the downloading time

I've ripped this module from tengine you can grab it here (I haven't
checked the new release of tengine for updates to this module) Joshua
was kind enough to help me get the examples in the README:

https://github.com/perusio/nginx-http-concat

--- appa


From piotr.sikora at frickle.com  Thu Jan 12 18:53:26 2012
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Thu, 12 Jan 2012 19:53:26 +0100
Subject: [ANNOUNCE] Tengine-1.2.2
In-Reply-To: <87aa5syd32.wl%appa@perusio.net>
References: <CAOKiq2C8czdC8+XBOoxuB04cq+g-jVooz2owt=5WLPDnx2x1OQ@mail.gmail.com>
	<C45999C885E7CA46B0E5847C6D1F0CC8DE2F9E@nasanexd01b.na.qualcomm.com>
	<87aa5syd32.wl%appa@perusio.net>
Message-ID: <677D8226993F4E9BA56899B090F61324@Desktop>

Hi,

> https://github.com/perusio/nginx-http-concat

It doesn't have license.

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


From piotr.sikora at frickle.com  Thu Jan 12 18:56:07 2012
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Thu, 12 Jan 2012 19:56:07 +0100
Subject: Programmatic access to Nginx fcgi/proxy cache?
In-Reply-To: <87d3aoygoe.wl%appa@perusio.net>
References: <OFD9E71ECE.515D64F4-ON80257983.00537F94-80257983.0053CC64@lionresources.com>
	<87d3aoygoe.wl%appa@perusio.net>
Message-ID: <64C9547719F2449EB2E21039892068DC@Desktop>

Hi,

> There's also a 3rd party module for cache purging. AFAIK doesn't
> support wildcards:
>
> https://github.com/FRiCKLE/ngx_cache_purge

It doesn't, but because of that it's the only one of those mentioned that 
works at scale.

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


From alabdelouahab at gmail.com  Thu Jan 12 20:15:02 2012
From: alabdelouahab at gmail.com (aliane abdelouahab)
Date: Thu, 12 Jan 2012 21:15:02 +0100
Subject: nginx for windows 64 (not the cygwin)
In-Reply-To: <201201121803.21728.ne@vbart.ru>
References: <dfa4c8912a3627d4b9526974c1f7c09b.NginxMailingListEnglish@forum.nginx.org>
	<201201121735.59770.ne@vbart.ru>
	<CAP4kiWoT8R98GL7qodt1eE0RmU0bk9W_MSoDbZVotg9KqMFJWw@mail.gmail.com>
	<201201121803.21728.ne@vbart.ru>
Message-ID: <CAP4kiWqXh+u3xBEoTAyToJv=cixJPD5W4NMrNuMZOzYt5qorcQ@mail.gmail.com>

so the default module are not included?
extras for example like ssl

2012/1/12, Valentin V. Bartenev <ne at vbart.ru>:
> On Thursday 12 January 2012 17:40:20 aliane abdelouahab wrote:
>> i run under 8 giga ram, so i just wanted to test the limits of a server
>> running under 64 bits, my thesis is to make a webserver using tornado
>> (also
>> not officialy supported on windows) and as load balancer and reverse proxy
>> i'll use nginx (as suggested in tornado documentation)
>
> http://nginx.org/en/docs/windows.html - this limitations relate to all
> windows
> builds, not only 32-bit. Don't expect high performance or more than 1024
> simultaneous connections.
>
>> and what about the 32 bits build, does it contain the basic modules +
>> extras?
>
> What "extras" do you mean exactly? Some of nginx modules don't work properly
> on
> windows 7, and, probably, most of 3'rd party modules even won't compile.
>
> wbr, Valentin V. Bartenev
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>


From appa at perusio.net  Thu Jan 12 20:27:36 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 12 Jan 2012 20:27:36 +0000
Subject: [ANNOUNCE] Tengine-1.2.2
In-Reply-To: <677D8226993F4E9BA56899B090F61324@Desktop>
References: <CAOKiq2C8czdC8+XBOoxuB04cq+g-jVooz2owt=5WLPDnx2x1OQ@mail.gmail.com>
	<C45999C885E7CA46B0E5847C6D1F0CC8DE2F9E@nasanexd01b.na.qualcomm.com>
	<87aa5syd32.wl%appa@perusio.net>
	<677D8226993F4E9BA56899B090F61324@Desktop>
Message-ID: <877h0wy8cn.wl%appa@perusio.net>

On 12 Jan 2012 18h53 WET, piotr.sikora at frickle.com wrote:

> Hi,
>
>> https://github.com/perusio/nginx-http-concat
>
> It doesn't have license.

Yes indeed. Fixed now.

Thanks,

--- appa


From ne at vbart.ru  Thu Jan 12 21:41:05 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Fri, 13 Jan 2012 01:41:05 +0400
Subject: nginx for windows 64 (not the cygwin)
In-Reply-To: <CAP4kiWqXh+u3xBEoTAyToJv=cixJPD5W4NMrNuMZOzYt5qorcQ@mail.gmail.com>
References: <dfa4c8912a3627d4b9526974c1f7c09b.NginxMailingListEnglish@forum.nginx.org>
	<201201121803.21728.ne@vbart.ru>
	<CAP4kiWqXh+u3xBEoTAyToJv=cixJPD5W4NMrNuMZOzYt5qorcQ@mail.gmail.com>
Message-ID: <201201130141.05751.ne@vbart.ru>

On Friday 13 January 2012 00:15:02 aliane abdelouahab wrote:
> so the default module are not included?
> extras for example like ssl
> 

You may download and check it yourself by using the "-V" option.

TLS SNI support enabled
configure arguments:
--with-cc=cl
--builddir=objs.msvc8
--with-debug
--prefix=
--conf-path=conf/nginx.conf
--pid-path=logs/nginx.pid
--http-log-path=logs/access.log
--error-log-path=logs/error.log
--sbin-path=nginx.exe
--http-client-body-temp-path=temp/client_body_temp
--http-proxy-temp-path=temp/proxy_temp
--http-fastcgi-temp-path=temp/fastcgi_temp
--with-cc-opt=-DFD_SETSIZE=1024
--with-pcre=objs.msvc8/lib/pcre-8.12
--with-zlib=objs.msvc8/lib/zlib-1.2.5
--with-select_module
--with-http_realip_module
--with-http_addition_module
--with-http_sub_module
--with-http_dav_module
--with-http_stub_status_module
--with-http_flv_module
--with-http_mp4_module
--with-http_gzip_static_module
--with-http_random_index_module
--with-http_secure_link_module
--with-mail
--with-openssl=objs.msvc8/lib/openssl-1.0.0e
--with-openssl-opt=enable-tlsext
--with-http_ssl_module
--with-mail_ssl_module
--with-ipv6

wbr, Valentin V. Bartenev


From nginx-forum at nginx.us  Thu Jan 12 22:14:27 2012
From: nginx-forum at nginx.us (johnny)
Date: Thu, 12 Jan 2012 17:14:27 -0500
Subject: How to respond correctly to a POST request for a static file?
In-Reply-To: <1326393766.10162.19.camel@portable-evil>
References: <1326393766.10162.19.camel@portable-evil>
Message-ID: <28994e7a5254b00484a7f260cf60ad05.NginxMailingListEnglish@forum.nginx.org>

Thanks, but I'm getting a 404, not a 405. Using the solution you linked
to would require all 404s to be 200, which is not a good idea. Any was
to limit this to just index.html?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221059,221133#msg-221133


From ne at vbart.ru  Thu Jan 12 22:29:50 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Fri, 13 Jan 2012 02:29:50 +0400
Subject: help with ssl config nginx -> thin -> rails
In-Reply-To: <CAEDYxyzNzYaOwz5De6rNr+hT1OtOtW4O94=Z917zvYkEiSruMQ@mail.gmail.com>
References: <CAEDYxyzNzYaOwz5De6rNr+hT1OtOtW4O94=Z917zvYkEiSruMQ@mail.gmail.com>
Message-ID: <201201130229.50669.ne@vbart.ru>

On Thursday 12 January 2012 22:30:43 Karl Johnson wrote:
[...]
> ^.+.(jpg|jpeg|gif|png|ico|css|zip|tgz|gz|rar|bz2|doc|xls|exe|ppt|txt|tar|mi
> d|midi|wav|bmp|rtf|ttf|woff|svg|eot)$ {
>         root   /var/www/dev/public/;
>                 expires           10y;
>                 add_header Cache-Control public;
>         }
> location ~ ^(/javascripts/.*)$ {
>         root   /var/www/dev/public/;
>                 expires           10y;
>                 add_header Cache-Control public;
>         }

Why did you put regexp captures everywhere and don't use them?

> 
>  if (-f $request_filename/index.html) {
>         rewrite (.*) $1/index.html break;
>       }
>       if (-f $request_filename.html) {
>         rewrite (.*) $1.html break;
>       }
>       if (!-f $request_filename) {
>         proxy_pass http://developmentmode;
>         break;
>       }

It's ugly. Please read:
http://wiki.nginx.org/IfIsEvil
http://wiki.nginx.org/Pitfalls

..and use the "try_files" directive:
http://nginx.org/en/docs/http/ngx_http_core_module.html#try_files

[...]
> }
> }
> server {
> listen 443;
> server_name govenga.com www.govenga.com;
[...]
> proxy_pass https://devmode-secure;
[...]

and above:

> 
> upstream devmode-secure{
> server 127.0.0.1:443;
> }
> 

...looks like you set up a loopback proxy pass to nginx itself.


wbr, Valentin V. Bartenev


From nginx-forum at nginx.us  Fri Jan 13 04:36:41 2012
From: nginx-forum at nginx.us (takigama)
Date: Thu, 12 Jan 2012 23:36:41 -0500
Subject: alias, regex and try_files am i doing it wrong, or is this a bug?
Message-ID: <24e6fcd6bfe45e3c8fcf79a0a2d0cbf8.NginxMailingListEnglish@forum.nginx.org>

Ok, i was trying to do drupal as a subdirectory on my site, but i've hit
a problem and I think its a bug. Im using 10.04 ubuntu, the nginx ppa
with 1.0.11 nginx. 

If i have the following config's and i try to retrieve the file
http://mysite/drupal/somefile:

location ~ /drupal(.*) {
    index index.php;
    alias /local_config/$1;
    try_files $uri @drupalrewrite;
}

The file nginx is apparently trying retrieve
"/local_config//somefile/drupal/somefile" rather then
"/local_config//somefile", which leads me to believe $uri isnt being set
correctly by nginx.

If instead, i have this as a config:

location /drupal/ {
        index index.php;
        alias /local_config/;
        try_files $uri @drupalrewrite;
}

it does the right thing, and try files will attempt to get
"/local_config/somefile", so im really confused as to why it happens.
Either im doing something wrong or in the first example, $uri gets set
incorrectly? some small chunks of debug output to show what i mean:

>From the first config:
2012/01/13 04:03:18 [debug] 22504#0: *4 http request line: "GET
/drupal/somefile HTTP/1.1"
2012/01/13 04:03:18 [debug] 22504#0: *4 http uri: "/drupal/somefile"
2012/01/13 04:03:18 [debug] 22504#0: *4 http args: ""
...
2012/01/13 04:03:18 [debug] 22504#0: *4 using configuration
"/drupal(.*)"
...
2012/01/13 04:03:18 [debug] 22504#0: *4 trying to use file:
"/drupal/somefile" "/local_config//somefile/drupal/somefile" <-- path
looks incorrect?


>From the second config:
2012/01/13 04:26:58 [debug] 23019#0: *1 http request line: "GET
/drupal/somefile HTTP/1.1"
2012/01/13 04:26:58 [debug] 23019#0: *1 http uri: "/drupal/somefile"
...
2012/01/13 04:26:58 [debug] 23019#0: *1 using configuration "/drupal/"
...
2012/01/13 04:26:58 [debug] 23019#0: *1 trying to use file: "somefile"
"/local_config/somefile" <--- correct

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221138,221138#msg-221138


From albert at nn.iij4u.or.jp  Fri Jan 13 05:46:08 2012
From: albert at nn.iij4u.or.jp (shin fukuda)
Date: Fri, 13 Jan 2012 14:46:08 +0900
Subject: http requests consolidation
In-Reply-To: <20120112132130.GD67687@mdounin.ru>
References: <20120112191526.607f450a02c1d8ac828c684b@nn.iij4u.or.jp>
	<20120112132130.GD67687@mdounin.ru>
Message-ID: <20120113144608.a0f27c287ef241804661c893@nn.iij4u.or.jp>

Hi Maxim.

Thx!!

On Thu, 12 Jan 2012 17:21:30 +0400
Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
> 
> On Thu, Jan 12, 2012 at 07:15:26PM +0900, shin fukuda wrote:
> 
> > Hi
> > 
> > We use Nginx as http proxy cache. 
> > Is Nginx supporting http requests consolidation ?
> > Or Is there a solution for that?
> 
> http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_cache_lock
> 
> Maxim Dounin
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


-- 
shin fukuda <albert at nn.iij4u.or.jp>


From nginx-forum at nginx.us  Fri Jan 13 06:55:38 2012
From: nginx-forum at nginx.us (locojohn)
Date: Fri, 13 Jan 2012 01:55:38 -0500
Subject: Any way to deal with clients breaking the limit zone
In-Reply-To: <20120112161129.GI67687@mdounin.ru>
References: <20120112161129.GI67687@mdounin.ru>
Message-ID: <7a35df337affc7dcc7541b2a18785329.NginxMailingListEnglish@forum.nginx.org>

Hello Maxim,

I am surely aware of how server requests are matched in nginx.   Thanks
for the info though.

The thing is, I have multiple virtual hosts configured all with
server_name defined, of course.   Then there's default fallback server
configured as this:

# default fallback server
        server {
                listen 80 default_server;
                server_name  _;
                return       444;
        }

After reading your reply I went to check my error logs and the strange
thing is that error messages go both into one of my virtual host
server's error log with correct data (remote address, server name,
etc.), but also go to nginx default (?) error_log with entries that I
reported?   How come, if the virtual host server being accessed using
the slowhttptest program is clearly "testsite" with its own error_log,
yet the connection limit error message is reported in two error logs?

slowhttptest -c 1000 -r 1000 -X -u http://testsite

results:

/var/log/nginx/testsite/error_log:

2012/01/12 18:10:34 [error] 10508#0: *15985 limiting connections by zone
"addr", client: 217.24.78.177, server: testsite, request: "GET /
HTTP/1.1", host: "devel.ahlerstoday.com", referrer:
"http://code.google.com/p/slowhttptest/"
2012/01/12 18:10:35 [error] 10508#0: *15989 limiting connections by zone
"addr", client: 217.24.78.177, server: testsite, request: "GET /
HTTP/1.1", host: "devel.ahlerstoday.com", referrer:
"http://code.google.com/p/slowhttptest/"

/var/log/nginx/error_log:

2012/01/12 18:10:34 [error] 10509#0: *12134 limiting connections by zone
"addr", client: 127.0.0.1, server: , request: "GET / HTTP/1.0", host:
"testsite", referrer: "http://code.google.com/p/slowhttptest/"
2012/01/12 18:10:35 [error] 10509#0: *12136 limiting connections by zone
"addr", client: 127.0.0.1, server: , request: "GET / HTTP/1.0", host:
"testsite", referrer: "http://code.google.com/p/slowhttptest/"


Could it have anything to do with the fact that limit_zone and
limit_conn are defined on http level?   Yet, I think the behaviour is
not fully correct.

I would appreciate your feedback,

Andrejs

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,203013,221143#msg-221143


From nginx-forum at nginx.us  Fri Jan 13 06:57:37 2012
From: nginx-forum at nginx.us (locojohn)
Date: Fri, 13 Jan 2012 01:57:37 -0500
Subject: Any way to deal with clients breaking the limit zone
In-Reply-To: <7a35df337affc7dcc7541b2a18785329.NginxMailingListEnglish@forum.nginx.org>
References: <20120112161129.GI67687@mdounin.ru>
	<7a35df337affc7dcc7541b2a18785329.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <a03e9314ffc91ee94b356431ed77f0f1.NginxMailingListEnglish@forum.nginx.org>

Note: Unfortunately, I cannot edit my messages here.  In both error log
entries, Host: is the same:  "testsite".

Andrejs

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,203013,221144#msg-221144


From nginx-forum at nginx.us  Fri Jan 13 08:57:09 2012
From: nginx-forum at nginx.us (voidandany)
Date: Fri, 13 Jan 2012 03:57:09 -0500
Subject: Protect a specific php file
In-Reply-To: <87ehv4yib1.wl%appa@perusio.net>
References: <87ehv4yib1.wl%appa@perusio.net>
Message-ID: <b78c07ac0dc6d173ae04887f6d104a05.NginxMailingListEnglish@forum.nginx.org>

It was already duplicate like this :

location ^~ /myapp/index.php {
auth_basic "Private Section";
auth_basic_user_file $document_root/.htpasswd;
location ~ \.php$ {
include /etc/nginx/fastcgi_params;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
fastcgi_pass 127.0.0.1:9000;
}
}

The inside location ~ \.php$ is silly because so I remove it like you
said, but no change.

By the way the problem is not when I use the complete url
(http://www.domain.fr/myapp/index.php) but when I don't add the
index.php page (http://www.domain.fr/myapp/)

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221094,221150#msg-221150


From agentzh at gmail.com  Fri Jan 13 09:11:01 2012
From: agentzh at gmail.com (agentzh)
Date: Fri, 13 Jan 2012 17:11:01 +0800
Subject: [ANN] ngx_echo module v0.37 released!
In-Reply-To: <CAB4Tn6PgdpZePDK_y4gygnDbrGqcHbSVQVYcJopiEaHPY2bs0g@mail.gmail.com>
References: <CAB4Tn6PgdpZePDK_y4gygnDbrGqcHbSVQVYcJopiEaHPY2bs0g@mail.gmail.com>
Message-ID: <CAB4Tn6PB1W7=oNs=qjNwtaCHpKFRH-LVZQp1C7S=x+FOLio8JA@mail.gmail.com>

Hi, folks!

I'm happy to announce the v0.37 release of the ngx_echo module, which
includes lots of bugfixes in the last 6 months.

You can get the release tarball from the download page below

? https://github.com/agentzh/echo-nginx-module/tags

Here's the complete change log for this release (compared to the last
release, v0.36):

* bugfix: data truncation might occur with echo_after_body in
not-so-good networks (we should be prepared for NGX_AGAIN returned by
downstream output filters). thanks Kindy Lin for reporting it.

* bugfix: we no longer check sync buffers for subrequests because it
is incorrect.

* bugfix: fixed a bug when sending out response headers: we did not
take into account the NGX_ERROR error code returned by
ngx_http_send_header.

* bugfix: we did not work with HEAD http requests before.

* bugfix: now we carefully eliminate empty flush buffers in
echo_after_body to work around a long-standing bug in the standard
HttpGzipModule.

* bugfix: we might send empty chain link in echo_after_body because it
may trigger the infamous "the http output chain is empty" alert in
error.log when the standard HttpSsiModule is disabled. thanks Sparsh
Gupta.

* bugfix: we did not set subrequest's Content-Length request headers
which could cause problems in the backends.

* bugfix: echo_exec + named locations might cause weird issues and now
we explicitly clear all the modules' contexts before calling
ngx_http_named_location.

* bugfix: echo_exec might hang when running after echo_sleep (or other
I/O interruption calls): we should have called
ngx_http_finalize_request on NGX_DONE to decrement r->main->count
anyway.

* bugfix: there was a memory issue in both echo_sleep and
echo_blocking_sleep: we should not pass ngx_str_t strings to atof()
which expects C strings.

* bugfix: some users report that this module cannot be compiled with
Nginx 1.0.x on their systems due to ngx_time_update (as in github
issue #7). this is a blind attemp to fix it because we could not
reproduce it on our side.

* bugfix: fixed places in the source code that we did not check null
pointers returned by the memory allocator.

You can also view the HTML-version of the change log here (which may
look better):

? ?http://wiki.nginx.org/HttpEchoModule#v0.37

The ngx_echo module wraps lots of Nginx internal APIs for streaming
input and output, parallel/sequential subrequests, timers and
sleeping, as well as various meta data accessing. You can read the
full documentation on the following wiki page:

? http://wiki.nginx.org/HttpEchoModule

and you can always get the latest source code from GitHub:

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

The ngx_echo module is included and enabled by default in our
ngx_openresty bundle:

? http://openresty.org

Happy echoing!
-agentzh


From nginx-forum at nginx.us  Fri Jan 13 09:52:25 2012
From: nginx-forum at nginx.us (bigplum)
Date: Fri, 13 Jan 2012 04:52:25 -0500
Subject: test-nginx doesn't work
Message-ID: <980096e872cfbf4c5be7fca64b373f40.NginxMailingListEnglish@forum.nginx.org>

I've put test-nginx into a test directory,  but it shows can't locate
socket.pm. Do I need to install TEST::NGINX into perl lib directory?

# ll test
??? 60
drwxr-xr-x 5 root  root  4096 2012-01-01 08:00 ./
drwxr-xr-x 5 simon simon 4096 2012-01-13 17:31 ../
-rw-r--r-- 1 root  root  8018 2012-01-01 08:00 Changes
-rw-r--r-- 1 root  root   298 2012-01-01 08:00 .gitignore
drwxr-xr-x 4 root  root  4096 2012-01-01 08:00 inc/
drwxr-xr-x 3 root  root  4096 2012-01-01 08:00 lib/
-rw-r--r-- 1 root  root   758 2012-01-01 08:00 Makefile.PL
-rw-r--r-- 1 root  root   749 2012-01-01 08:00 MANIFEST
-rw-r--r-- 1 root  root   303 2012-01-01 08:00 MANIFEST.SKIP
-rw-r--r-- 1 root  root   867 2012-01-01 08:00 META.yml
-rw-r--r-- 1 root  root  9851 2012-01-01 08:00 README
drwxr-xr-x 2 root  root  4096 2012-01-01 08:00 t/


# prove -r t/apply_moves.t 
t/apply_moves.t .. Can't locate Test/Nginx/Socket.pm in @INC (@INC
contains: /etc/perl /usr/local/lib/perl/5.10.1
/usr/local/share/perl/5.10.1 /usr/lib/perl5 /usr/share/perl5
/usr/lib/perl/5.10 /usr/share/perl/5.10 /usr/local/lib/site_perl .) at
t/apply_moves.t line 2.
BEGIN failed--compilation aborted at t/apply_moves.t line 2.
t/apply_moves.t .. Dubious, test returned 2 (wstat 512, 0x200)
No subtests run 

Test Summary Report
-------------------
t/apply_moves.t (Wstat: 512 Tests: 0 Failed: 0)
  Non-zero exit status: 2
  Parse errors: No plan found in TAP output
Files=1, Tests=0,  0 wallclock secs ( 0.02 usr +  0.01 sys =  0.03 CPU)
Result: FAIL

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221154,221154#msg-221154


From piotr.sikora at frickle.com  Fri Jan 13 09:57:41 2012
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Fri, 13 Jan 2012 10:57:41 +0100
Subject: test-nginx doesn't work
In-Reply-To: <980096e872cfbf4c5be7fca64b373f40.NginxMailingListEnglish@forum.nginx.org>
References: <980096e872cfbf4c5be7fca64b373f40.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <F0EEDB20F3804F028974928BD04911F6@Desktop>

Hi,

> # prove -r t/apply_moves.t 

prove -I lib -r t

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


From zhuzhaoyuan at gmail.com  Fri Jan 13 10:04:11 2012
From: zhuzhaoyuan at gmail.com (Joshua Zhu)
Date: Fri, 13 Jan 2012 18:04:11 +0800
Subject: [PATCH 2 of 2] add support for setting worker processes and CPU
	affinity automatically
In-Reply-To: <201201122153.56491.ne@vbart.ru>
References: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>
	<201201122153.56491.ne@vbart.ru>
Message-ID: <CAOKiq2AXXb4tQHvDxmcxfeSQDo4ZZDmma2LC5f7ct2ekUKwY3w@mail.gmail.com>

Hi,

On Fri, Jan 13, 2012 at 1:53 AM, Valentin V. Bartenev <ne at vbart.ru> wrote:

> On Thursday 12 January 2012 19:37:15 Joshua Zhu wrote:
>

[snip]

Thank you very much for the review.

 You use "ngx_ncpu" global variable and for Linux you set it by
> sysconf(_SC_NPROCESSORS_ONLN).
>
> On freeBSD and MacOS systems it set by sysctls(hw.ncpu), and it is equal
> to 1 on others.
>

Setting CPU affinity is available on Linux systems only. Please correct me
if I'm wrong.


> Which means that your "auto" settings are very inaccurate. Sometimes the
> "worker_processes" directive will be equal to 2 * cores or more, sometimes
> it will be equal to 1, and if you're lucky, then it will be equal to real
> CPU cores.
>

This is mostly caused by the hyper-threading feature of some CPUs. I don't
think this is an issue if hyper-threading is turned on. Sure we need to
find a better way to detect the number of CPUs on a system.


> So, I don't think that this patch makes life easier. In fact, it creates a
> false sense of optimal worker_processes setting. Even if ngx_ncpu would be
> always equal to real hardware cores number, it wouldn't be optimal for
> every
> use cases.
>

It's not 'optimal', but 'automatic'. This patch still supports setting
worker processes and CPU affinities mannually as what the standard Nginx
does ;)

Regards,

-- 
Joshua Zhu
Senior Software Engineer
Server Platforms Team at Taobao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/ad96f235/attachment-0001.html>

From maxim at nginx.com  Fri Jan 13 10:15:33 2012
From: maxim at nginx.com (Maxim Konovalov)
Date: Fri, 13 Jan 2012 14:15:33 +0400
Subject: [PATCH 2 of 2] add support for setting worker processes and CPU
	affinity automatically
In-Reply-To: <CAOKiq2AXXb4tQHvDxmcxfeSQDo4ZZDmma2LC5f7ct2ekUKwY3w@mail.gmail.com>
References: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>
	<201201122153.56491.ne@vbart.ru>
	<CAOKiq2AXXb4tQHvDxmcxfeSQDo4ZZDmma2LC5f7ct2ekUKwY3w@mail.gmail.com>
Message-ID: <4F100445.6030508@nginx.com>

> [snip]
>
> Thank you very much for the review.
>
>     You use "ngx_ncpu" global variable and for Linux you set it by
>     sysconf(_SC_NPROCESSORS_ONLN).
>
>     On freeBSD and MacOS systems it set by sysctls(hw.ncpu), and it
>     is equal
>     to 1 on others.
>
> Setting CPU affinity is available on Linux systems only. Please
> correct me if I'm wrong.
>
That's wrong.

Just some start points:

FreeBSD: cpuset_setaffinity(2)
Solaris: probably better to start from processor_bind(2) man page.
HP/UP: pset_bind(2), psrset(1m)
AIX: bindprocessor()

Sure other systems with more or less advances *MP support have a
similar functionality.

-- 
Maxim Konovalov
+7 (910) 4293178
http://nginx.com/


From piotr.sikora at frickle.com  Fri Jan 13 10:27:32 2012
From: piotr.sikora at frickle.com (Piotr Sikora)
Date: Fri, 13 Jan 2012 11:27:32 +0100
Subject: [PATCH 2 of 2] add support for setting worker processes and CPU
	affinity automatically
In-Reply-To: <4F100445.6030508@nginx.com>
References: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>
	<201201122153.56491.ne@vbart.ru>
	<CAOKiq2AXXb4tQHvDxmcxfeSQDo4ZZDmma2LC5f7ct2ekUKwY3w@mail.gmail.com>
	<4F100445.6030508@nginx.com>
Message-ID: <337DD761F4634F2EB7B8E5D1624E1105@Desktop>

Hi,

>> Setting CPU affinity is available on Linux systems only. Please
>> correct me if I'm wrong.
>
> That's wrong.

I believe that Joshua meant that workers CPU affinity _in nginx_
works only on Linux right now (which is true).

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


From ne at vbart.ru  Fri Jan 13 10:33:52 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Fri, 13 Jan 2012 14:33:52 +0400
Subject: [PATCH 2 of 2] add support for setting worker processes and CPU
	affinity automatically
In-Reply-To: <CAOKiq2AXXb4tQHvDxmcxfeSQDo4ZZDmma2LC5f7ct2ekUKwY3w@mail.gmail.com>
References: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>
	<201201122153.56491.ne@vbart.ru>
	<CAOKiq2AXXb4tQHvDxmcxfeSQDo4ZZDmma2LC5f7ct2ekUKwY3w@mail.gmail.com>
Message-ID: <201201131433.53292.ne@vbart.ru>

On Friday 13 January 2012 14:04:11 Joshua Zhu wrote:
> Hi,
> 
> On Fri, Jan 13, 2012 at 1:53 AM, Valentin V. Bartenev <ne at vbart.ru> wrote:
> > On Thursday 12 January 2012 19:37:15 Joshua Zhu wrote:
> [snip]
> 
> Thank you very much for the review.
> 
>  You use "ngx_ncpu" global variable and for Linux you set it by
> 
> > sysconf(_SC_NPROCESSORS_ONLN).
> > 
> > On freeBSD and MacOS systems it set by sysctls(hw.ncpu), and it is equal
> > to 1 on others.
> 
> Setting CPU affinity is available on Linux systems only. Please correct me
> if I'm wrong.
> 

You're right.

But I generally said about "ngx_ncpu", that also affects "auto" setting in your 
patch for the "worker_processes" directive.

> > Which means that your "auto" settings are very inaccurate. Sometimes the
> > "worker_processes" directive will be equal to 2 * cores or more,
> > sometimes it will be equal to 1, and if you're lucky, then it will be
> > equal to real CPU cores.
> 
> This is mostly caused by the hyper-threading feature of some CPUs. I don't
> think this is an issue if hyper-threading is turned on. Sure we need to
> find a better way to detect the number of CPUs on a system.
> 

Yeap, we need to find a better way. And I don't think the solution will be simple 
and it is worth to be implemented in a web server. 

> > So, I don't think that this patch makes life easier. In fact, it creates
> > a false sense of optimal worker_processes setting. Even if ngx_ncpu
> > would be always equal to real hardware cores number, it wouldn't be
> > optimal for every
> > use cases.
> 
> It's not 'optimal', but 'automatic'. This patch still supports setting
> worker processes and CPU affinities mannually as what the standard Nginx
> does ;)

So, when you need some automation for configuration and deployment, the better 
solution would be using proper software for that task, isn't it? Like Puppet,
for example.

wbr, Valentin V. Bartenev


From mdounin at mdounin.ru  Fri Jan 13 10:37:08 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 13 Jan 2012 14:37:08 +0400
Subject: Any way to deal with clients breaking the limit zone
In-Reply-To: <7a35df337affc7dcc7541b2a18785329.NginxMailingListEnglish@forum.nginx.org>
References: <20120112161129.GI67687@mdounin.ru>
	<7a35df337affc7dcc7541b2a18785329.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120113103708.GY67687@mdounin.ru>

Hello!

On Fri, Jan 13, 2012 at 01:55:38AM -0500, locojohn wrote:

> Hello Maxim,
> 
> I am surely aware of how server requests are matched in nginx.   Thanks
> for the info though.
> 
> The thing is, I have multiple virtual hosts configured all with
> server_name defined, of course.   Then there's default fallback server
> configured as this:
> 
> # default fallback server
>         server {
>                 listen 80 default_server;
>                 server_name  _;
>                 return       444;
>         }
> 
> After reading your reply I went to check my error logs and the strange
> thing is that error messages go both into one of my virtual host
> server's error log with correct data (remote address, server name,
> etc.), but also go to nginx default (?) error_log with entries that I
> reported?   How come, if the virtual host server being accessed using
> the slowhttptest program is clearly "testsite" with its own error_log,
> yet the connection limit error message is reported in two error logs?
> 
> slowhttptest -c 1000 -r 1000 -X -u http://testsite
> 
> results:
> 
> /var/log/nginx/testsite/error_log:
> 
> 2012/01/12 18:10:34 [error] 10508#0: *15985 limiting connections by zone
> "addr", client: 217.24.78.177, server: testsite, request: "GET /
> HTTP/1.1", host: "devel.ahlerstoday.com", referrer:
> "http://code.google.com/p/slowhttptest/"
> 2012/01/12 18:10:35 [error] 10508#0: *15989 limiting connections by zone
> "addr", client: 217.24.78.177, server: testsite, request: "GET /
> HTTP/1.1", host: "devel.ahlerstoday.com", referrer:
> "http://code.google.com/p/slowhttptest/"

Note HTTP/1.1.  It's likely to be original request.

> 
> /var/log/nginx/error_log:
> 
> 2012/01/12 18:10:34 [error] 10509#0: *12134 limiting connections by zone
> "addr", client: 127.0.0.1, server: , request: "GET / HTTP/1.0", host:
> "testsite", referrer: "http://code.google.com/p/slowhttptest/"
> 2012/01/12 18:10:35 [error] 10509#0: *12136 limiting connections by zone
> "addr", client: 127.0.0.1, server: , request: "GET / HTTP/1.0", host:
> "testsite", referrer: "http://code.google.com/p/slowhttptest/"

Note HTTP/1.0 here.  It looks like requests here appear after 
proxy_pass to 127.0.0.1 in your original "testsite" server.

The fact the request isn't matched to the same testsite server 
suggests that things happen on different socket, i.e. you use 
listen directive in the server block in question (the one without 
server_name) either on different port or explicitly on 127.0.0.1.

> Could it have anything to do with the fact that limit_zone and
> limit_conn are defined on http level?   Yet, I think the behaviour is
> not fully correct.

Behaviour seems correct for me.

Maxim Dounin


From agentzh at gmail.com  Fri Jan 13 10:57:11 2012
From: agentzh at gmail.com (agentzh)
Date: Fri, 13 Jan 2012 18:57:11 +0800
Subject: test-nginx doesn't work
In-Reply-To: <980096e872cfbf4c5be7fca64b373f40.NginxMailingListEnglish@forum.nginx.org>
References: <980096e872cfbf4c5be7fca64b373f40.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CAB4Tn6Nr6maukYvdM768dprDHWKcKDe-q4txy5-X2j+Atg9hUg@mail.gmail.com>

On Fri, Jan 13, 2012 at 5:52 PM, bigplum <nginx-forum at nginx.us> wrote:
> I've put test-nginx into a test directory, ?but it shows can't locate
> socket.pm. Do I need to install TEST::NGINX into perl lib directory?
>

You need to

1. install Test::Nginx into your perl installation by running the
following commands on shell:

    perl Makefile.PL
    make
    make install

In case that you have multiple perl installations in your system,
please ensure that you're using the right perl.

2. specify the location of Test::Nginx's lib/ directory in your
"prove" command line (just as Piotr Sikora has pointed out), that is

    prove -I/path/to/Test-Nginx/lib -r t

Regards,
-agentzh


From nginx-forum at nginx.us  Fri Jan 13 11:13:28 2012
From: nginx-forum at nginx.us (voidandany)
Date: Fri, 13 Jan 2012 06:13:28 -0500
Subject: Protect a specific php file
In-Reply-To: <b78c07ac0dc6d173ae04887f6d104a05.NginxMailingListEnglish@forum.nginx.org>
References: <87ehv4yib1.wl%appa@perusio.net>
	<b78c07ac0dc6d173ae04887f6d104a05.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <8b1ff087c6b7d626e6efe0e21e636f69.NginxMailingListEnglish@forum.nginx.org>

After some test I can give more detail.

In fact my app isn't directly in a subdirectory of my root, but in a
subdir of a subdir of the root.


root = /var/www/domain.fr;
myapp = /var/www/domain.fr/test/myapp


If I put it directly in a subdir of root it work


Here is my real conf :

server {
        listen       80;
        server_name  www.domain.fr;
        root         /var/www/domain.fr ;
        access_log      /var/log/nginx/access.log ;
        location / {
                deny all;
        }

        location /test/myapp {
                index  index.php;
                location ^~ /test/myapp/index.php {
                        auth_basic            "Section privee";
                        auth_basic_user_file 
$document_root/test/myapp/.htpasswd;
                        include         /etc/nginx/fastcgi_params;
                        fastcgi_param   SCRIPT_FILENAME 
$document_root$fastcgi_script_name;
                        fastcgi_pass    127.0.0.1:9000;
                }

                location ~ \.php$ {
                        include         /etc/nginx/fastcgi_params;
                        fastcgi_param   SCRIPT_FILENAME 
$document_root$fastcgi_script_name;
                        fastcgi_pass    127.0.0.1:9000;
                }
        }
}


With : http://www.domain.fr/test/myapp/index.php, password asked
With : http://www.domain.fr/test/myapp, no password asked, php file
downloaded

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221094,221165#msg-221165


From mdounin at mdounin.ru  Fri Jan 13 12:52:18 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 13 Jan 2012 16:52:18 +0400
Subject: alias, regex and try_files am i doing it wrong, or is this a bug?
In-Reply-To: <24e6fcd6bfe45e3c8fcf79a0a2d0cbf8.NginxMailingListEnglish@forum.nginx.org>
References: <24e6fcd6bfe45e3c8fcf79a0a2d0cbf8.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120113125218.GC67687@mdounin.ru>

Hello!

On Thu, Jan 12, 2012 at 11:36:41PM -0500, takigama wrote:

> Ok, i was trying to do drupal as a subdirectory on my site, but i've hit
> a problem and I think its a bug. Im using 10.04 ubuntu, the nginx ppa
> with 1.0.11 nginx. 
> 
> If i have the following config's and i try to retrieve the file
> http://mysite/drupal/somefile:
> 
> location ~ /drupal(.*) {
>     index index.php;
>     alias /local_config/$1;
>     try_files $uri @drupalrewrite;
> }
> 
> The file nginx is apparently trying retrieve
> "/local_config//somefile/drupal/somefile" rather then
> "/local_config//somefile", which leads me to believe $uri isnt being set
> correctly by nginx.

Yes, this looks like a bug for me.

(Internally, the problem is a bit complex, and I event tend to 
think that behaviour of try_files with alias in a regexp location 
is inherently undefined, but the example above should certainly 
work as expected.)

> If instead, i have this as a config:
> 
> location /drupal/ {
>         index index.php;
>         alias /local_config/;
>         try_files $uri @drupalrewrite;
> }
> 
> it does the right thing, and try files will attempt to get
> "/local_config/somefile", so im really confused as to why it happens.

This works as expected.

The difference is that in a regexp location the "alias" directive 
specifies full path to a file to use, and try_files isn't 
currently able to handle this.

Maxim Dounin


From martinloy.uy at gmail.com  Fri Jan 13 12:55:00 2012
From: martinloy.uy at gmail.com (Martin Loy)
Date: Fri, 13 Jan 2012 10:55:00 -0200
Subject: mod_aclr2
In-Reply-To: <10B30FF4-6B7B-4032-BBAF-495AF0C0E8EE@nginx.com>
References: <ED8F5676-8A10-4F60-94B8-C09812A4A954@nginx.com>
	<87sjjmxs2o.wl%appa@perusio.net>
	<CAHbGeW6rZpAX0dMXp6ovHbUnXxWx5_eUBbVHMbtfbpMAPR_oSw@mail.gmail.com>
	<10B30FF4-6B7B-4032-BBAF-495AF0C0E8EE@nginx.com>
Message-ID: <CAHbGeW7vqP1_yK-y6CqKJzcE94BUhS8mORr1kgoyj2peESZrUQ@mail.gmail.com>

Andrey

thanks for your response and great nice work!!

Regards

M

On Wed, Jan 11, 2012 at 4:09 PM, Andrey Belov <defan at nginx.com> wrote:

> Martin,
>
> On Jan 11, 2012, at 7:31 PM, Martin Loy wrote:
>
> Would you explain the main difference between using your module and using
> nginx as a reverse proxy and caching static content ?
>
>
> In some situations, mod_aclr2 helps to avoid excessive disk I/O,
> especially for big static responses.
>
> When nginx is used as a local reverse proxy and proxy_buffering is
> enabled, every upstream reponse which can not fit into proxy buffers
> may be temporarily saved on disk:
>
> http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_buffering
> http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_temp_path
>
> For example, we have to process the request for 700mb video file,
> without mod_aclr2:
>
> 1) nginx sends the request to apache
> 2) apache sends 700mb file back to nginx
> 3) nginx saves received file on disk (proxy_temp_path) and
> starts sending it back to the client
>
> and with mod_aclr2:
>
> 1) nginx sends the request to apache
> 2) apache sends a few bytes header with X-Accel-Redirect back to nginx
> 3) nginx starts sending the file directly from disk
>
> You save your time and your disk at step #2.
>
> Of course, you also may use proxy_cache along with mod_aclr2,
> but if your disks are fast enough, probably you won't need
> to do any caching at all.
>
> Hope this helps.
>
>
>
> Regards
>
> M
>
> On Wed, Jan 11, 2012 at 11:54 AM, Ant?nio P. P. Almeida <appa at perusio.net>wrote:
>
>> On 11 Jan 2012 13h17 WET, defan at nginx.com wrote:
>>
>> > Hello!
>> >
>> > Ever wondered how to set up NGINX as a local proxy for Apache 2.x
>> > with minimum efforts?
>>
>> Nice work Andrey. Opens up Nginx as a possibility for shared hosting
>> (keep using .htaccess) and voids any resistance people might have trying
>> Nginx out. It's sooo easy now :)
>>
>> Thanks,
>> --- appa
>>
>> _______________________________________________
>> nginx mailing list
>> nginx at nginx.org
>> http://mailman.nginx.org/mailman/listinfo/nginx
>>
>
>
>
> --
> *Nunca hubo un amigo que hiciese un favor a un enano, ni un enemigo que
> le hiciese un mal, que no se viese recompensado por entero.*
>  _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
>
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>



-- 
*Nunca hubo un amigo que hiciese un favor a un enano, ni un enemigo que le
hiciese un mal, que no se viese recompensado por entero.*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/615c8118/attachment.html>

From zhuzhaoyuan at gmail.com  Fri Jan 13 13:08:25 2012
From: zhuzhaoyuan at gmail.com (Joshua Zhu)
Date: Fri, 13 Jan 2012 21:08:25 +0800
Subject: [PATCH 2 of 2] add support for setting worker processes and CPU
	affinity automatically
In-Reply-To: <201201131433.53292.ne@vbart.ru>
References: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>
	<201201122153.56491.ne@vbart.ru>
	<CAOKiq2AXXb4tQHvDxmcxfeSQDo4ZZDmma2LC5f7ct2ekUKwY3w@mail.gmail.com>
	<201201131433.53292.ne@vbart.ru>
Message-ID: <CAOKiq2ADiejbL4SF7DDTqrMHntpf0tD2R1A_-fxJWNQs0O+j_g@mail.gmail.com>

Hi,

On Fri, Jan 13, 2012 at 6:33 PM, Valentin V. Bartenev <ne at vbart.ru> wrote:

> On Friday 13 January 2012 14:04:11 Joshua Zhu wrote:
>

[snip]


> So, when you need some automation for configuration and deployment, the
> better
> solution would be using proper software for that task, isn't it? Like
> Puppet,
> for example.
>
>
OK.
If this patch would not be accepted, could you please also review the first
one?
Thank you.

Regards,

-- 
Joshua Zhu
Senior Software Engineer
Server Platforms Team at Taobao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/f130682d/attachment.html>

From maxim at nginx.com  Fri Jan 13 13:18:51 2012
From: maxim at nginx.com (Maxim Konovalov)
Date: Fri, 13 Jan 2012 17:18:51 +0400
Subject: [PATCH 2 of 2] add support for setting worker processes and CPU
	affinity automatically
In-Reply-To: <337DD761F4634F2EB7B8E5D1624E1105@Desktop>
References: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>
	<201201122153.56491.ne@vbart.ru>
	<CAOKiq2AXXb4tQHvDxmcxfeSQDo4ZZDmma2LC5f7ct2ekUKwY3w@mail.gmail.com>
	<4F100445.6030508@nginx.com>
	<337DD761F4634F2EB7B8E5D1624E1105@Desktop>
Message-ID: <4F102F3B.5050506@nginx.com>

On 1/13/12 2:27 PM, Piotr Sikora wrote:
> Hi,
>
>>> Setting CPU affinity is available on Linux systems only. Please
>>> correct me if I'm wrong.
>>
>> That's wrong.
>
> I believe that Joshua meant that workers CPU affinity _in nginx_
> works only on Linux right now (which is true).
>
Ah, I see, thanks.

I just wonder how useful this option in real environment.

-- 
Maxim Konovalov
+7 (910) 4293178
http://nginx.com/


From nginx-forum at nginx.us  Fri Jan 13 13:36:20 2012
From: nginx-forum at nginx.us (double)
Date: Fri, 13 Jan 2012 08:36:20 -0500
Subject: Limit robots
Message-ID: <6300b8a589bb337813f7df86d393d332.NginxMailingListEnglish@forum.nginx.org>

Hello,

Is there a chance to limit robots to 1 request per second?
The down below does not work:  [emerg] "limit_req" directive is not
allowed here

Thanks a lot
Markus


http {
    limit_req_zone      $http_user_agent  zone=useragenttrack:1m 
rate=1r/s;
    server {
        listen          80;
        location / {
            if ($http_user_agent ~* "[a-z]bot[^a-z]") {
                limit_req   zone=useragenttrack  burst=100 nodelay;
            }
        }
    }
}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221162,221162#msg-221162


From nginx-forum at nginx.us  Fri Jan 13 13:44:13 2012
From: nginx-forum at nginx.us (andres-m77)
Date: Fri, 13 Jan 2012 08:44:13 -0500
Subject: help with replacing chars "_" to "." in a $variable
Message-ID: <7770f2f8745407086920bd36c17248d1.NginxMailingListEnglish@forum.nginx.org>

I would appreciate your help

I have to do the following request

http://subdomain_domain_extension_tdl.domain.com/

Where .domain.com must be deleted and subdomain_domain_extension_tdl
must be converted to subdomain.domain.extension.tdl


I did this:

        set $subdomain "";

        if ($host ~* "^(.+)\.domain\.com$") {
            set $subdomain $1;
            break;
        }

          resolver 8.8.8.8;
          proxy_pass        http://$subdomain;
          proxy_set_header  X-Real-IP  $remote_addr;


Then i get $subdomain as: subdomain_domain_extension_tdl but i need to
transform it into subdomain.domain.extension.tdl

Could you please help me determining how could i do this? i would like
to serve as proxy in this way:


If you want to get inside any blocked website like www.facebook.com be
able to do it throug www_facebook_com.domain.com

Also i know that perhaps sometimes subdomain_domain_extension_tdl could
be domain_extension_tdl, 


Thanks in advance!
Andr?s

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221176,221176#msg-221176


From appa at perusio.net  Fri Jan 13 13:50:48 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Fri, 13 Jan 2012 13:50:48 +0000
Subject: Limit robots
In-Reply-To: <6300b8a589bb337813f7df86d393d332.NginxMailingListEnglish@forum.nginx.org>
References: <6300b8a589bb337813f7df86d393d332.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <874nvzyamf.wl%appa@perusio.net>

On 13 Jan 2012 13h36 WET, nginx-forum at nginx.us wrote:

> Hello,
>
> Is there a chance to limit robots to 1 request per second?
> The down below does not work:  [emerg] "limit_req" directive is not
> allowed here

Yes. Try:

At the *http* level:

limit_req_zone $http_user_agent zone=useragenttrack:1m rate=1r/s; 

map $http_user_agent $is_bot {
    default 0;
    ~[a-z]bot[^a-z] 1;
}

At the server level:

location / {
    error_page 418 @bots;

    if ($is_bot) {
      return 418;  
    }
    ...
}

location @bots {
   limit_req zone=useragenttrack burst=100 nodelay;  
   ...
}

Cf: http://wiki.nginx.org/HttpLimitReqModule

--- appa


From martinloy.uy at gmail.com  Fri Jan 13 13:57:30 2012
From: martinloy.uy at gmail.com (Martin Loy)
Date: Fri, 13 Jan 2012 11:57:30 -0200
Subject: Programmatic access to Nginx fcgi/proxy cache?
In-Reply-To: <64C9547719F2449EB2E21039892068DC@Desktop>
References: <OFD9E71ECE.515D64F4-ON80257983.00537F94-80257983.0053CC64@lionresources.com>
	<87d3aoygoe.wl%appa@perusio.net>
	<64C9547719F2449EB2E21039892068DC@Desktop>
Message-ID: <CAHbGeW6f0xq6AyeZxGjcRb=p-5aMWdFP5Y7sBVUTSdQ3M5EZ3A@mail.gmail.com>

Ben

Why not use memcached/redis/etc as cache backend instead of file based
cache, nginx works with them great and you just need to populate the cache
at application level :)

http://wiki.nginx.org/HttpMemcachedModule
http://wiki.nginx.org/HttpRedis

Regards

M


On Thu, Jan 12, 2012 at 4:56 PM, Piotr Sikora <piotr.sikora at frickle.com>wrote:

> Hi,
>
>
>  There's also a 3rd party module for cache purging. AFAIK doesn't
>> support wildcards:
>>
>> https://github.com/FRiCKLE/**ngx_cache_purge<https://github.com/FRiCKLE/ngx_cache_purge>
>>
>
> It doesn't, but because of that it's the only one of those mentioned that
> works at scale.
>
> Best regards,
> Piotr Sikora < piotr.sikora at frickle.com >
>
>
> ______________________________**_________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/**mailman/listinfo/nginx<http://mailman.nginx.org/mailman/listinfo/nginx>
>



-- 
*Nunca hubo un amigo que hiciese un favor a un enano, ni un enemigo que le
hiciese un mal, que no se viese recompensado por entero.*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/8eaa408c/attachment-0001.html>

From mdounin at mdounin.ru  Fri Jan 13 14:06:15 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 13 Jan 2012 18:06:15 +0400
Subject: [PATCH 2 of 2] add support for setting worker processes and CPU
	affinity automatically
In-Reply-To: <CAOKiq2ADiejbL4SF7DDTqrMHntpf0tD2R1A_-fxJWNQs0O+j_g@mail.gmail.com>
References: <CAOKiq2DRGbNbwR0=vZmpqq57_2c_rkVjS_KwPtrzW=E_B+x_xA@mail.gmail.com>
	<201201122153.56491.ne@vbart.ru>
	<CAOKiq2AXXb4tQHvDxmcxfeSQDo4ZZDmma2LC5f7ct2ekUKwY3w@mail.gmail.com>
	<201201131433.53292.ne@vbart.ru>
	<CAOKiq2ADiejbL4SF7DDTqrMHntpf0tD2R1A_-fxJWNQs0O+j_g@mail.gmail.com>
Message-ID: <20120113140615.GH67687@mdounin.ru>

Hello!

On Fri, Jan 13, 2012 at 09:08:25PM +0800, Joshua Zhu wrote:

> Hi,
> 
> On Fri, Jan 13, 2012 at 6:33 PM, Valentin V. Bartenev <ne at vbart.ru> wrote:
> 
> > On Friday 13 January 2012 14:04:11 Joshua Zhu wrote:
> >
> 
> [snip]
> 
> 
> > So, when you need some automation for configuration and deployment, the
> > better
> > solution would be using proper software for that task, isn't it? Like
> > Puppet,
> > for example.
> >
> >
> OK.
> If this patch would not be accepted,

I wouldn't say it won't be accepted at all, I think we need 
something like this, but probably not by default.

> could you please also review the first one?

I'm currently looking into it.

Maxim Dounin


From xpoke07 at hotmail.com  Fri Jan 13 14:30:35 2012
From: xpoke07 at hotmail.com (tavo astuet)
Date: Fri, 13 Jan 2012 08:30:35 -0600
Subject: nginx Digest, Vol 27, Issue 33
In-Reply-To: <mailman.3575.1326463072.3032.nginx@nginx.org>
References: <mailman.3575.1326463072.3032.nginx@nginx.org>
Message-ID: <BLU123-W190C18FB18A19A0D84A1FBA19C0@phx.gbl>


alguien habla castellano??hello, which is the number of processes to host a site in nginx?You can have a server on site, which serves to re-send requests within the local network??

> From: nginx-request at nginx.org
> Subject: nginx Digest, Vol 27, Issue 33
> To: nginx at nginx.org
> Date: Fri, 13 Jan 2012 13:57:52 +0000
> 
> Send nginx mailing list submissions to
> 	nginx at nginx.org
> 
> To subscribe or unsubscribe via the World Wide Web, visit
> 	http://mailman.nginx.org/mailman/listinfo/nginx
> or, via email, send a message with subject or body 'help' to
> 	nginx-request at nginx.org
> 
> You can reach the person managing the list at
> 	nginx-owner at nginx.org
> 
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of nginx digest..."
> 
> 
> Today's Topics:
> 
>    1. Re: alias, regex and try_files am i doing it wrong, or is
>       this a bug? (Maxim Dounin)
>    2. Re: mod_aclr2 (Martin Loy)
>    3. Re: [PATCH 2 of 2] add support for setting worker processes
>       and CPU	affinity automatically (Joshua Zhu)
>    4. Re: [PATCH 2 of 2] add support for setting worker processes
>       and CPU	affinity automatically (Maxim Konovalov)
>    5. Limit robots (double)
>    6. help with replacing chars "_" to "." in a $variable (andres-m77)
>    7. Re: Limit robots (Ant?nio P. P. Almeida)
>    8. Re: Programmatic access to Nginx fcgi/proxy cache? (Martin Loy)
> 
> 
> ----------------------------------------------------------------------
> 
> Message: 1
> Date: Fri, 13 Jan 2012 16:52:18 +0400
> From: Maxim Dounin <mdounin at mdounin.ru>
> To: nginx at nginx.org
> Subject: Re: alias, regex and try_files am i doing it wrong, or is
> 	this a bug?
> Message-ID: <20120113125218.GC67687 at mdounin.ru>
> Content-Type: text/plain; charset=us-ascii
> 
> Hello!
> 
> On Thu, Jan 12, 2012 at 11:36:41PM -0500, takigama wrote:
> 
> > Ok, i was trying to do drupal as a subdirectory on my site, but i've hit
> > a problem and I think its a bug. Im using 10.04 ubuntu, the nginx ppa
> > with 1.0.11 nginx. 
> > 
> > If i have the following config's and i try to retrieve the file
> > http://mysite/drupal/somefile:
> > 
> > location ~ /drupal(.*) {
> >     index index.php;
> >     alias /local_config/$1;
> >     try_files $uri @drupalrewrite;
> > }
> > 
> > The file nginx is apparently trying retrieve
> > "/local_config//somefile/drupal/somefile" rather then
> > "/local_config//somefile", which leads me to believe $uri isnt being set
> > correctly by nginx.
> 
> Yes, this looks like a bug for me.
> 
> (Internally, the problem is a bit complex, and I event tend to 
> think that behaviour of try_files with alias in a regexp location 
> is inherently undefined, but the example above should certainly 
> work as expected.)
> 
> > If instead, i have this as a config:
> > 
> > location /drupal/ {
> >         index index.php;
> >         alias /local_config/;
> >         try_files $uri @drupalrewrite;
> > }
> > 
> > it does the right thing, and try files will attempt to get
> > "/local_config/somefile", so im really confused as to why it happens.
> 
> This works as expected.
> 
> The difference is that in a regexp location the "alias" directive 
> specifies full path to a file to use, and try_files isn't 
> currently able to handle this.
> 
> Maxim Dounin
> 
> 
> 
> ------------------------------
> 
> Message: 2
> Date: Fri, 13 Jan 2012 10:55:00 -0200
> From: Martin Loy <martinloy.uy at gmail.com>
> To: nginx at nginx.org
> Subject: Re: mod_aclr2
> Message-ID:
> 	<CAHbGeW7vqP1_yK-y6CqKJzcE94BUhS8mORr1kgoyj2peESZrUQ at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
> 
> Andrey
> 
> thanks for your response and great nice work!!
> 
> Regards
> 
> M
> 
> On Wed, Jan 11, 2012 at 4:09 PM, Andrey Belov <defan at nginx.com> wrote:
> 
> > Martin,
> >
> > On Jan 11, 2012, at 7:31 PM, Martin Loy wrote:
> >
> > Would you explain the main difference between using your module and using
> > nginx as a reverse proxy and caching static content ?
> >
> >
> > In some situations, mod_aclr2 helps to avoid excessive disk I/O,
> > especially for big static responses.
> >
> > When nginx is used as a local reverse proxy and proxy_buffering is
> > enabled, every upstream reponse which can not fit into proxy buffers
> > may be temporarily saved on disk:
> >
> > http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_buffering
> > http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_temp_path
> >
> > For example, we have to process the request for 700mb video file,
> > without mod_aclr2:
> >
> > 1) nginx sends the request to apache
> > 2) apache sends 700mb file back to nginx
> > 3) nginx saves received file on disk (proxy_temp_path) and
> > starts sending it back to the client
> >
> > and with mod_aclr2:
> >
> > 1) nginx sends the request to apache
> > 2) apache sends a few bytes header with X-Accel-Redirect back to nginx
> > 3) nginx starts sending the file directly from disk
> >
> > You save your time and your disk at step #2.
> >
> > Of course, you also may use proxy_cache along with mod_aclr2,
> > but if your disks are fast enough, probably you won't need
> > to do any caching at all.
> >
> > Hope this helps.
> >
> >
> >
> > Regards
> >
> > M
> >
> > On Wed, Jan 11, 2012 at 11:54 AM, Ant?nio P. P. Almeida <appa at perusio.net>wrote:
> >
> >> On 11 Jan 2012 13h17 WET, defan at nginx.com wrote:
> >>
> >> > Hello!
> >> >
> >> > Ever wondered how to set up NGINX as a local proxy for Apache 2.x
> >> > with minimum efforts?
> >>
> >> Nice work Andrey. Opens up Nginx as a possibility for shared hosting
> >> (keep using .htaccess) and voids any resistance people might have trying
> >> Nginx out. It's sooo easy now :)
> >>
> >> Thanks,
> >> --- appa
> >>
> >> _______________________________________________
> >> nginx mailing list
> >> nginx at nginx.org
> >> http://mailman.nginx.org/mailman/listinfo/nginx
> >>
> >
> >
> >
> > --
> > *Nunca hubo un amigo que hiciese un favor a un enano, ni un enemigo que
> > le hiciese un mal, que no se viese recompensado por entero.*
> >  _______________________________________________
> > nginx mailing list
> > nginx at nginx.org
> > http://mailman.nginx.org/mailman/listinfo/nginx
> >
> >
> >
> > _______________________________________________
> > nginx mailing list
> > nginx at nginx.org
> > http://mailman.nginx.org/mailman/listinfo/nginx
> >
> 
> 
> 
> -- 
> *Nunca hubo un amigo que hiciese un favor a un enano, ni un enemigo que le
> hiciese un mal, que no se viese recompensado por entero.*
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/615c8118/attachment-0001.html>
> 
> ------------------------------
> 
> Message: 3
> Date: Fri, 13 Jan 2012 21:08:25 +0800
> From: Joshua Zhu <zhuzhaoyuan at gmail.com>
> To: nginx at nginx.org
> Subject: Re: [PATCH 2 of 2] add support for setting worker processes
> 	and CPU	affinity automatically
> Message-ID:
> 	<CAOKiq2ADiejbL4SF7DDTqrMHntpf0tD2R1A_-fxJWNQs0O+j_g at mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
> 
> Hi,
> 
> On Fri, Jan 13, 2012 at 6:33 PM, Valentin V. Bartenev <ne at vbart.ru> wrote:
> 
> > On Friday 13 January 2012 14:04:11 Joshua Zhu wrote:
> >
> 
> [snip]
> 
> 
> > So, when you need some automation for configuration and deployment, the
> > better
> > solution would be using proper software for that task, isn't it? Like
> > Puppet,
> > for example.
> >
> >
> OK.
> If this patch would not be accepted, could you please also review the first
> one?
> Thank you.
> 
> Regards,
> 
> -- 
> Joshua Zhu
> Senior Software Engineer
> Server Platforms Team at Taobao
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/f130682d/attachment-0001.html>
> 
> ------------------------------
> 
> Message: 4
> Date: Fri, 13 Jan 2012 17:18:51 +0400
> From: Maxim Konovalov <maxim at nginx.com>
> To: nginx at nginx.org
> Subject: Re: [PATCH 2 of 2] add support for setting worker processes
> 	and CPU	affinity automatically
> Message-ID: <4F102F3B.5050506 at nginx.com>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
> 
> On 1/13/12 2:27 PM, Piotr Sikora wrote:
> > Hi,
> >
> >>> Setting CPU affinity is available on Linux systems only. Please
> >>> correct me if I'm wrong.
> >>
> >> That's wrong.
> >
> > I believe that Joshua meant that workers CPU affinity _in nginx_
> > works only on Linux right now (which is true).
> >
> Ah, I see, thanks.
> 
> I just wonder how useful this option in real environment.
> 
> -- 
> Maxim Konovalov
> +7 (910) 4293178
> http://nginx.com/
> 
> 
> 
> ------------------------------
> 
> Message: 5
> Date: Fri, 13 Jan 2012 08:36:20 -0500
> From: "double" <nginx-forum at nginx.us>
> To: nginx at nginx.org
> Subject: Limit robots
> Message-ID:
> 	<6300b8a589bb337813f7df86d393d332.NginxMailingListEnglish at forum.nginx.org>
> 	
> Content-Type: text/plain; charset=UTF-8
> 
> Hello,
> 
> Is there a chance to limit robots to 1 request per second?
> The down below does not work:  [emerg] "limit_req" directive is not
> allowed here
> 
> Thanks a lot
> Markus
> 
> 
> http {
>     limit_req_zone      $http_user_agent  zone=useragenttrack:1m 
> rate=1r/s;
>     server {
>         listen          80;
>         location / {
>             if ($http_user_agent ~* "[a-z]bot[^a-z]") {
>                 limit_req   zone=useragenttrack  burst=100 nodelay;
>             }
>         }
>     }
> }
> 
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221162,221162#msg-221162
> 
> 
> 
> ------------------------------
> 
> Message: 6
> Date: Fri, 13 Jan 2012 08:44:13 -0500
> From: "andres-m77" <nginx-forum at nginx.us>
> To: nginx at nginx.org
> Subject: help with replacing chars "_" to "." in a $variable
> Message-ID:
> 	<7770f2f8745407086920bd36c17248d1.NginxMailingListEnglish at forum.nginx.org>
> 	
> Content-Type: text/plain; charset=UTF-8
> 
> I would appreciate your help
> 
> I have to do the following request
> 
> http://subdomain_domain_extension_tdl.domain.com/
> 
> Where .domain.com must be deleted and subdomain_domain_extension_tdl
> must be converted to subdomain.domain.extension.tdl
> 
> 
> I did this:
> 
>         set $subdomain "";
> 
>         if ($host ~* "^(.+)\.domain\.com$") {
>             set $subdomain $1;
>             break;
>         }
> 
>           resolver 8.8.8.8;
>           proxy_pass        http://$subdomain;
>           proxy_set_header  X-Real-IP  $remote_addr;
> 
> 
> Then i get $subdomain as: subdomain_domain_extension_tdl but i need to
> transform it into subdomain.domain.extension.tdl
> 
> Could you please help me determining how could i do this? i would like
> to serve as proxy in this way:
> 
> 
> If you want to get inside any blocked website like www.facebook.com be
> able to do it throug www_facebook_com.domain.com
> 
> Also i know that perhaps sometimes subdomain_domain_extension_tdl could
> be domain_extension_tdl, 
> 
> 
> Thanks in advance!
> Andr?s
> 
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221176,221176#msg-221176
> 
> 
> 
> ------------------------------
> 
> Message: 7
> Date: Fri, 13 Jan 2012 13:50:48 +0000
> From: Ant?nio P. P. Almeida <appa at perusio.net>
> To: nginx at nginx.org
> Subject: Re: Limit robots
> Message-ID: <874nvzyamf.wl%appa at perusio.net>
> Content-Type: text/plain; charset=US-ASCII
> 
> On 13 Jan 2012 13h36 WET, nginx-forum at nginx.us wrote:
> 
> > Hello,
> >
> > Is there a chance to limit robots to 1 request per second?
> > The down below does not work:  [emerg] "limit_req" directive is not
> > allowed here
> 
> Yes. Try:
> 
> At the *http* level:
> 
> limit_req_zone $http_user_agent zone=useragenttrack:1m rate=1r/s; 
> 
> map $http_user_agent $is_bot {
>     default 0;
>     ~[a-z]bot[^a-z] 1;
> }
> 
> At the server level:
> 
> location / {
>     error_page 418 @bots;
> 
>     if ($is_bot) {
>       return 418;  
>     }
>     ...
> }
> 
> location @bots {
>    limit_req zone=useragenttrack burst=100 nodelay;  
>    ...
> }
> 
> Cf: http://wiki.nginx.org/HttpLimitReqModule
> 
> --- appa
> 
> 
> 
> ------------------------------
> 
> Message: 8
> Date: Fri, 13 Jan 2012 11:57:30 -0200
> From: Martin Loy <martinloy.uy at gmail.com>
> To: nginx at nginx.org
> Subject: Re: Programmatic access to Nginx fcgi/proxy cache?
> Message-ID:
> 	<CAHbGeW6f0xq6AyeZxGjcRb=p-5aMWdFP5Y7sBVUTSdQ3M5EZ3A at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
> 
> Ben
> 
> Why not use memcached/redis/etc as cache backend instead of file based
> cache, nginx works with them great and you just need to populate the cache
> at application level :)
> 
> http://wiki.nginx.org/HttpMemcachedModule
> http://wiki.nginx.org/HttpRedis
> 
> Regards
> 
> M
> 
> 
> On Thu, Jan 12, 2012 at 4:56 PM, Piotr Sikora <piotr.sikora at frickle.com>wrote:
> 
> > Hi,
> >
> >
> >  There's also a 3rd party module for cache purging. AFAIK doesn't
> >> support wildcards:
> >>
> >> https://github.com/FRiCKLE/**ngx_cache_purge<https://github.com/FRiCKLE/ngx_cache_purge>
> >>
> >
> > It doesn't, but because of that it's the only one of those mentioned that
> > works at scale.
> >
> > Best regards,
> > Piotr Sikora < piotr.sikora at frickle.com >
> >
> >
> > ______________________________**_________________
> > nginx mailing list
> > nginx at nginx.org
> > http://mailman.nginx.org/**mailman/listinfo/nginx<http://mailman.nginx.org/mailman/listinfo/nginx>
> >
> 
> 
> 
> -- 
> *Nunca hubo un amigo que hiciese un favor a un enano, ni un enemigo que le
> hiciese un mal, que no se viese recompensado por entero.*
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/8eaa408c/attachment.html>
> 
> ------------------------------
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
> 
> End of nginx Digest, Vol 27, Issue 33
> *************************************
 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/550168dc/attachment-0001.html>

From karljohn at gmail.com  Fri Jan 13 15:42:42 2012
From: karljohn at gmail.com (Karl Johnson)
Date: Fri, 13 Jan 2012 10:42:42 -0500
Subject: help with ssl config nginx -> thin -> rails
In-Reply-To: <201201130229.50669.ne@vbart.ru>
References: <CAEDYxyzNzYaOwz5De6rNr+hT1OtOtW4O94=Z917zvYkEiSruMQ@mail.gmail.com>
	<201201130229.50669.ne@vbart.ru>
Message-ID: <CAEDYxywv-WBO8_JRv0DixOk3T0VU0vXjOsNom3XK=hGLvaFOcQ@mail.gmail.com>

Ignore all of the regex's for now, they were left over from a previous
developer who configured the instance. I'll remove them.

My biggest concern though is around what the address(s) should be for
devmode-secure.
What should that look like in the config if the IP plus 443 causes it to
loopback to nginx?

Thanks for your help.
Karl

On Thu, Jan 12, 2012 at 5:29 PM, Valentin V. Bartenev <ne at vbart.ru> wrote:

> On Thursday 12 January 2012 22:30:43 Karl Johnson wrote:
> [...]
> >
> ^.+.(jpg|jpeg|gif|png|ico|css|zip|tgz|gz|rar|bz2|doc|xls|exe|ppt|txt|tar|mi
> > d|midi|wav|bmp|rtf|ttf|woff|svg|eot)$ {
> >         root   /var/www/dev/public/;
> >                 expires           10y;
> >                 add_header Cache-Control public;
> >         }
> > location ~ ^(/javascripts/.*)$ {
> >         root   /var/www/dev/public/;
> >                 expires           10y;
> >                 add_header Cache-Control public;
> >         }
>
> Why did you put regexp captures everywhere and don't use them?
>
> >
> >  if (-f $request_filename/index.html) {
> >         rewrite (.*) $1/index.html break;
> >       }
> >       if (-f $request_filename.html) {
> >         rewrite (.*) $1.html break;
> >       }
> >       if (!-f $request_filename) {
> >         proxy_pass http://developmentmode;
> >         break;
> >       }
>
> It's ugly. Please read:
> http://wiki.nginx.org/IfIsEvil
> http://wiki.nginx.org/Pitfalls
>
> ..and use the "try_files" directive:
> http://nginx.org/en/docs/http/ngx_http_core_module.html#try_files
>
> [...]
> > }
> > }
> > server {
> > listen 443;
> > server_name govenga.com www.govenga.com;
> [...]
> > proxy_pass https://devmode-secure;
> [...]
>
> and above:
>
> >
> > upstream devmode-secure{
> > server 127.0.0.1:443;
> > }
> >
>
> ...looks like you set up a loopback proxy pass to nginx itself.
>
>
> wbr, Valentin V. Bartenev
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/bf91dfdc/attachment.html>

From brian at akins.org  Fri Jan 13 15:46:19 2012
From: brian at akins.org (Brian Akins)
Date: Fri, 13 Jan 2012 10:46:19 -0500
Subject: Programmatic access to Nginx fcgi/proxy cache?
In-Reply-To: <CAHbGeW6f0xq6AyeZxGjcRb=p-5aMWdFP5Y7sBVUTSdQ3M5EZ3A@mail.gmail.com>
References: <OFD9E71ECE.515D64F4-ON80257983.00537F94-80257983.0053CC64@lionresources.com>
	<87d3aoygoe.wl%appa@perusio.net>
	<64C9547719F2449EB2E21039892068DC@Desktop>
	<CAHbGeW6f0xq6AyeZxGjcRb=p-5aMWdFP5Y7sBVUTSdQ3M5EZ3A@mail.gmail.com>
Message-ID: <1D5A41E2-66BF-4602-82CD-46EC8E50B7A7@akins.org>


On Jan 13, 2012, at 8:57 AM, Martin Loy wrote:

> 
> http://wiki.nginx.org/HttpMemcachedModule

the memc module is better for this, at least in my experience:  http://wiki.nginx.org/HttpMemcModule

> http://wiki.nginx.org/HttpRedis

only does gets. redis2 module supports most redis commands: http://wiki.nginx.org/HttpRedis2Module

I had issues in the past with the app writing into the cache and nginx reading from it. Partially because of the numerous apps we run nginx in front of.  I like nginx doing the reading and the writing.

this has worked well: http://wiki.nginx.org/NginxHttpSRCacheModule
Or some simple lua using location captures.

if you used redis, for example, you could do something in lua to do the purging that took regexes fairly easily.

--Brian

From ne at vbart.ru  Fri Jan 13 16:43:20 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Fri, 13 Jan 2012 20:43:20 +0400
Subject: help with ssl config nginx -> thin -> rails
In-Reply-To: <CAEDYxywv-WBO8_JRv0DixOk3T0VU0vXjOsNom3XK=hGLvaFOcQ@mail.gmail.com>
References: <CAEDYxyzNzYaOwz5De6rNr+hT1OtOtW4O94=Z917zvYkEiSruMQ@mail.gmail.com>
	<201201130229.50669.ne@vbart.ru>
	<CAEDYxywv-WBO8_JRv0DixOk3T0VU0vXjOsNom3XK=hGLvaFOcQ@mail.gmail.com>
Message-ID: <201201132043.20832.ne@vbart.ru>

On Friday 13 January 2012 19:42:42 Karl Johnson wrote:
[...]
> My biggest concern though is around what the address(s) should be for
> devmode-secure.

As usual, address and port that your backend listening.

> What should that look like in the config if the IP plus 443 causes it to
> loopback to nginx?
> 

What do you want exactly? Use nginx for SSL and proxy https to http backend?

Or proxy https to https transparently?


wbr, Valentin V. Bartenev


From casey.jordan at jorsek.com  Fri Jan 13 17:47:32 2012
From: casey.jordan at jorsek.com (Casey Jordan)
Date: Fri, 13 Jan 2012 12:47:32 -0500
Subject: NGINX CSS optimizer module?
Message-ID: <CAPAQPqdzSUmvnEFPn3bbaCCE11WMNF2gDwRxpgwA30mQfSJgkg@mail.gmail.com>

Hi all,

We have been using NGINX for a year now with great results, so thanks to
everyone who has made this project so successful!

I am looking for a feature which I hope NGINX supports, and that is
optimization of serving CSS files, specifically around minimizing and
resolving @import statements server side.

For instance, to better manage the hundreds of css files we have, we
started breaking them up and using the @import directive. Obviously this
slows things down quite a bit because the browser has to make lots of
additional requests.

I am looking for a way to have NGINX automatically resolve all @imports so
when it servers a CSS file so it has no dependencies.

Any feedback would be much appreciated!

Thanks,

Casey

-- 
--
Casey Jordan
easyDITA a product of Jorsek LLC
"CaseyDJordan" on LinkedIn, Twitter & Facebook
(585) 348 7399
easydita.com


This message is intended only for the use of the Addressee(s) and may
contain information that is privileged, confidential, and/or exempt from
disclosure under applicable law.  If you are not the intended recipient,
please be advised that any disclosure  copying, distribution, or use of
the information contained herein is prohibited.  If you have received
this communication in error, please destroy all copies of the message,
whether in electronic or hard copy format, as well as attachments, and
immediately contact the sender by replying to this e-mail or by phone.
Thank you.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/a2967ecb/attachment.html>

From ne at vbart.ru  Fri Jan 13 18:08:09 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Fri, 13 Jan 2012 22:08:09 +0400
Subject: NGINX CSS optimizer module?
In-Reply-To: <CAPAQPqdzSUmvnEFPn3bbaCCE11WMNF2gDwRxpgwA30mQfSJgkg@mail.gmail.com>
References: <CAPAQPqdzSUmvnEFPn3bbaCCE11WMNF2gDwRxpgwA30mQfSJgkg@mail.gmail.com>
Message-ID: <201201132208.09158.ne@vbart.ru>

On Friday 13 January 2012 21:47:32 Casey Jordan wrote:
[...]
> 
> I am looking for a feature which I hope NGINX supports, and that is
> optimization of serving CSS files, specifically around minimizing and
> resolving @import statements server side.
> 
> For instance, to better manage the hundreds of css files we have, we
> started breaking them up and using the @import directive. Obviously this
> slows things down quite a bit because the browser has to make lots of
> additional requests.
> 
> I am looking for a way to have NGINX automatically resolve all @imports so
> when it servers a CSS file so it has no dependencies.
> 
> Any feedback would be much appreciated!
> 

Do you want to do such resource-intensive task on every request?

wbr, Valentin V. Bartenev


From casey.jordan at jorsek.com  Fri Jan 13 18:10:45 2012
From: casey.jordan at jorsek.com (Casey Jordan)
Date: Fri, 13 Jan 2012 13:10:45 -0500
Subject: NGINX CSS optimizer module?
In-Reply-To: <201201132208.09158.ne@vbart.ru>
References: <CAPAQPqdzSUmvnEFPn3bbaCCE11WMNF2gDwRxpgwA30mQfSJgkg@mail.gmail.com>
	<201201132208.09158.ne@vbart.ru>
Message-ID: <CAPAQPqd2O0Ldd91P6gc69zf7aXjiES6TNzh3mfRfsUJteWOBAA@mail.gmail.com>

Ideally, on production server it would cache the result and serve that on
all following requests.

However, on development servers we would need to disable caching of the
files so that we could see our changes take effect in real time.

Thanks,

Casey



On Fri, Jan 13, 2012 at 1:08 PM, Valentin V. Bartenev <ne at vbart.ru> wrote:

> On Friday 13 January 2012 21:47:32 Casey Jordan wrote:
> [...]
> >
> > I am looking for a feature which I hope NGINX supports, and that is
> > optimization of serving CSS files, specifically around minimizing and
> > resolving @import statements server side.
> >
> > For instance, to better manage the hundreds of css files we have, we
> > started breaking them up and using the @import directive. Obviously this
> > slows things down quite a bit because the browser has to make lots of
> > additional requests.
> >
> > I am looking for a way to have NGINX automatically resolve all @imports
> so
> > when it servers a CSS file so it has no dependencies.
> >
> > Any feedback would be much appreciated!
> >
>
> Do you want to do such resource-intensive task on every request?
>
> wbr, Valentin V. Bartenev
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>



-- 
--
Casey Jordan
easyDITA a product of Jorsek LLC
"CaseyDJordan" on LinkedIn, Twitter & Facebook
(585) 348 7399
easydita.com


This message is intended only for the use of the Addressee(s) and may
contain information that is privileged, confidential, and/or exempt from
disclosure under applicable law.  If you are not the intended recipient,
please be advised that any disclosure  copying, distribution, or use of
the information contained herein is prohibited.  If you have received
this communication in error, please destroy all copies of the message,
whether in electronic or hard copy format, as well as attachments, and
immediately contact the sender by replying to this e-mail or by phone.
Thank you.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/addfe2d7/attachment.html>

From mdounin at mdounin.ru  Fri Jan 13 18:16:51 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 13 Jan 2012 22:16:51 +0400
Subject: [PATCH 1 of 2] fix sched_setaffinity(2) call failure
In-Reply-To: <CAOKiq2D2hwCFy3+9wFeU8ZVxtHRgU3Nrj4-UafSuYhgn8AzCOw@mail.gmail.com>
References: <CAOKiq2D2hwCFy3+9wFeU8ZVxtHRgU3Nrj4-UafSuYhgn8AzCOw@mail.gmail.com>
Message-ID: <20120113181651.GM67687@mdounin.ru>

Hello!

On Thu, Jan 12, 2012 at 11:37:00PM +0800, Joshua Zhu wrote:

> Hi,
> 
> 'cpu_set_t' should be used instead of 'long' when calling
> sched_setaffinity(2). Otherwise it may fail on some Linux systems.
> Please see the attachment for more detail.

According to sched_setaffinity(2) at least in kernels >= 2.6.9 it 
should be ok to use any length as long as bits above kernel mask 
used aren't referenced.

I believe the real problem is that code uses 32 as cpusetsize, 
while this is the size in bytes.  Hence it references arbitrary 
memory and this causes EINVAL to be returned by glibc if there are 
bits referenced above kernel cpumask size, see 
sysdeps/unix/sysv/linux/sched_setaffinity.c in glibc's sources.

The following patch should be enough:

diff -r e1296af53cc0 src/os/unix/ngx_process_cycle.c
--- a/src/os/unix/ngx_process_cycle.c	Mon Dec 26 00:00:00 2011 +0400
+++ b/src/os/unix/ngx_process_cycle.c	Fri Jan 13 11:58:01 2012 -0500
@@ -914,7 +914,10 @@
         ngx_log_error(NGX_LOG_NOTICE, cycle->log, 0,
                       "sched_setaffinity(0x%08Xl)", cpu_affinity);
 
-        if (sched_setaffinity(0, 32, (cpu_set_t *) &cpu_affinity) == -1) {
+        if (sched_setaffinity(0, sizeof(cpu_affinity),
+                              (cpu_set_t *) &cpu_affinity)
+            == -1)
+        {
             ngx_log_error(NGX_LOG_ALERT, cycle->log, ngx_errno,
                           "sched_setaffinity(0x%08Xl) failed", cpu_affinity);
         }

Could you please test if it works for you?

It passes my limited testing, but I wasn't really able to 
reproduce the problem without modifying the sched_setaffinity() 
call in question to use bigger cpusetsize.

[...]

> diff -uprN nginx-1.1.12/src/core/nginx.c nginx-1.1.12-bugfix/src/core/nginx.c
> --- nginx-1.1.12/src/core/nginx.c	2011-11-16 04:35:42.000000000 +0800
> +++ nginx-1.1.12-bugfix/src/core/nginx.c	2012-01-12 11:06:00.000000000 +0800
> @@ -1245,15 +1245,15 @@ ngx_set_cpu_affinity(ngx_conf_t *cf, ngx
>      ngx_core_conf_t  *ccf = conf;
>  
>      u_char            ch;
> -    u_long           *mask;
> +    cpu_set_t        *mask;
>      ngx_str_t        *value;
> -    ngx_uint_t        i, n;
> +    ngx_uint_t        i, j, n;
>  
>      if (ccf->cpu_affinity) {
>          return "is duplicate";
>      }
>  
> -    mask = ngx_palloc(cf->pool, (cf->args->nelts - 1) * sizeof(long));
> +    mask = ngx_palloc(cf->pool, (cf->args->nelts - 1) * sizeof(cpu_set_t));
>      if (mask == NULL) {
>          return NGX_CONF_ERROR;
>      }

[...]

While I generally agree that using cpu_set_t and CPU_* macros 
explicitly would be better, I don't think we want to introduce 
linux-centric code here in platform-independent code.

Instead we probably want to leave this code generic enough to be 
useable with other scheduler interfaces, and convert the data 
somewhere near platform-specific whatever_setaffinity() calls if 
needed.  (Likely unneeded for the Linux with the above patch.)

Maxim Dounin


From francis at daoine.org  Fri Jan 13 18:29:34 2012
From: francis at daoine.org (Francis Daly)
Date: Fri, 13 Jan 2012 18:29:34 +0000
Subject: Protect a specific php file
In-Reply-To: <8b1ff087c6b7d626e6efe0e21e636f69.NginxMailingListEnglish@forum.nginx.org>
References: <87ehv4yib1.wl%appa@perusio.net>
	<b78c07ac0dc6d173ae04887f6d104a05.NginxMailingListEnglish@forum.nginx.org>
	<8b1ff087c6b7d626e6efe0e21e636f69.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120113182934.GA27443@craic.sysops.org>

On Fri, Jan 13, 2012 at 06:13:28AM -0500, voidandany wrote:

Hi there,

> With : http://www.domain.fr/test/myapp/index.php, password asked
> With : http://www.domain.fr/test/myapp, no password asked, php file
> downloaded

Your configuration looks like it should not result in what you report.

When I use a very similar config with 1.1.11, I do not see what you
report, and I do see what you expect.

With the following config:

==
    server {
        listen       8000;
        include fastcgi.conf;

        location / {
            deny all;
        }

        location /test/myapp {
            index index.php;
            location ^~ /test/myapp/index.php {
                auth_basic            "Section privee";
                auth_basic_user_file  $document_root/test/myapp/.htpasswd;
                fastcgi_pass  unix:php.sock;
            }
            location ~ \.php$ {
                fastcgi_pass  unix:php.sock;
            }
        }
    }

==

curl -i http://localhost:8000/test returns 403 (Forbidden)
curl -i http://localhost:8000/test/myapp returns 301 (Moved to /test/myapp/)
curl -i http://localhost:8000/test/myapp/ returns 401 (Unauthorized)
curl -i -u x:x http://localhost:8000/test/myapp/ returns 200 (output of index.php)
curl -i http://localhost:8000/test/myapp/index.php returns 401 (Unauthorized)
curl -i -u x:x http://localhost:8000/test/myapp/index.php returns 200 (output of index.php)
curl -i http://localhost:8000/test/myapp/other.php returns 200 (output of other.php)

The extra things I would point out are that you allow simple downloading
of all other content that begins with the location /test/myapp, which
includes /test/myapp2 (if that directory exists), and which also includes
/test/myapp/.htpasswd; and from the configuration shown, it's probably
more elegant to use "=" instead of "^~" in the nested location.

So, what's different between your test and mine?

Are there any other location{}s defined in your config? Did the browser
you were testing with have an empty cache, so that you saw the request
in access.log?

Good luck with it,

	f
-- 
Francis Daly        francis at daoine.org


From andrew at nginx.com  Fri Jan 13 19:43:28 2012
From: andrew at nginx.com (Andrew Alexeev)
Date: Fri, 13 Jan 2012 23:43:28 +0400
Subject: fastcgi_keep_conn consuming ports?
In-Reply-To: <20120108000339.GU67687@mdounin.ru>
References: <918068D0CE3C4A48880F53D3EF19861919630D1875@BRANDNETEXCH.Brandnet.local>
	<ca1192fc0d5e483409b37332d51f61f7.NginxMailingListEnglish@forum.nginx.org>
	<20120108000339.GU67687@mdounin.ru>
Message-ID: <FE4F5CE0-097D-4423-B77A-44F57DCD9CBC@nginx.com>

On Jan 8, 2012, at 4:03 AM, Maxim Dounin wrote:

> Hello!
> 
> On Sat, Jan 07, 2012 at 06:08:09PM -0500, locojohn wrote:
> 
>> Hello Maxim,
>> 
>> Is there any place where both directives are well documented?  I tried
>> wiki and the new http://nginx.org/en/docs/, but no luck so far..
> 
> Official documentation is expected to appear soon on nginx.org.  
> 
> Maxim Dounin

It's now here (done by Ruslan Ermilov)

http://nginx.org/en/docs/http/ngx_http_upstream_module.html#keepalive
http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_http_version
http://nginx.org/en/docs/http/ngx_http_fastcgi_module.html#fastcgi_keep_conn



From nginx-forum at nginx.us  Fri Jan 13 19:47:00 2012
From: nginx-forum at nginx.us (mikeyk)
Date: Fri, 13 Jan 2012 14:47:00 -0500
Subject: proxy_intercept_errors and logging
Message-ID: <3a36efae80ac4fa8adf082759636ac86.NginxMailingListEnglish@forum.nginx.org>

Hi there,

We have nginx proxying to a bunch of django upstreams.

When those upstreams throw an error, we use proxy_intercept_errors to
show a custom page. The problem is that this causes those 500s to be
logged into the global access_log, rather than the access_log specified
for that location. Here is a (reduced) example:

    # at the server 
    access_log /logs/nginx/access.log accesslogformat;
    proxy_intercept_errors on;
    location ~ ^/oauth.*$ {
        proxy_pass        http://appservers;

        access_log /logs/nginx/app.log applogformat;
    }

Is there a way to get errors to log to the app.log in this case, even
with proxy_intercept_errors to on? Right now we get all the 500s in the
access.log.

Thanks in advance!
Mike

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221201,221201#msg-221201


From varia at e-healthexpert.org  Fri Jan 13 19:51:41 2012
From: varia at e-healthexpert.org (Mark Alan)
Date: Fri, 13 Jan 2012 19:51:41 +0000
Subject: EGroupWare nginx config
Message-ID: <20120113195141.3ad9aa14@e-healthexpert.org>

In case anyone needs to setup EGroupWare http://www.egroupware.org/
over Debian 6 / Ubuntu 11.10 and Nginx 1.1.12, here is the config file
that we used:

# nginx - /etc/nginx/sites-available/egw
#
server {
	listen 80;
	server_name egw.localhost;
	root /usr/share/egroupware;
	index index.php;
	client_max_body_size 8M;
	rewrite ^/egroupware/(.*)$ /$1 last;
	try_files $uri $uri/ /index.php?$args;
	location ~ \.php$ {
		include /etc/nginx/fastcgi_params;
		fastcgi_pass unix:/tmp/php-fpm.socket;
		fastcgi_index index.php;
		fastcgi_param  SCRIPT_FILENAME
$document_root$fastcgi_script_name; }
}


M.


From varia at e-healthexpert.org  Fri Jan 13 20:01:24 2012
From: varia at e-healthexpert.org (Mark Alan)
Date: Fri, 13 Jan 2012 20:01:24 +0000
Subject: EGroupWare nginx config v.2 (without try_files)
Message-ID: <20120113200124.465801bd@e-healthexpert.org>

In case anyone needs to setup EGroupWare http://www.egroupware.org/
over Debian 6 / Ubuntu 11.10 and Nginx 1.1.12, here is the config file
that we used:

# nginx - /etc/nginx/sites-available/egw
#
server {
	listen 80;
	server_name egw.localhost;
	root /usr/share/egroupware;
	index index.php;
	client_max_body_size 8M;
	rewrite ^/egroupware/(.*)$ /$1 last;
	try_files $uri $uri/ /index.php?$args;
		location ~ ^(?<script>.+\.php) {
		include /etc/nginx/fastcgi_params;
		fastcgi_pass unix:/tmp/php-fpm.socket;
		fastcgi_index index.php;
		fastcgi_param SCRIPT_FILENAME $document_root$script;
		fastcgi_param SCRIPT_NAME $script;
	}
}


M.


From varia at e-healthexpert.org  Fri Jan 13 20:08:20 2012
From: varia at e-healthexpert.org (Mark Alan)
Date: Fri, 13 Jan 2012 20:08:20 +0000
Subject: Feng Office nginx config
Message-ID: <20120113200820.504f2c83@e-healthexpert.org>

In case anyone needs to setup Feng Office
http://www.fengoffice.com/web/features.php (it is a web-based groupware
package, including project management, client relationship management,
email calendar, billing, financing, etc.),
over Debian 6 / Ubuntu 11.10 and Nginx 1.1.12, here is the
config file that we used:

# nginx - /etc/nginx/sites-available/feng
#
server {
	listen 80;
	server_name feng.localhost;
	access_log /var/log/nginx/feng.access.log;
	error_log /var/log/nginx/feng.error.log;
	root /var/www/fengoffice;
	index index.php;
	client_max_body_size 8M;
	location ~ ^(?<script>.+\.php) {
		include /etc/nginx/fastcgi_params;
		fastcgi_pass unix:/tmp/php-fpm.socket;
		fastcgi_index index.php;
		fastcgi_param SCRIPT_FILENAME $document_root$script;
		fastcgi_param SCRIPT_NAME $script;
	}
}

M.


From alabdelouahab at gmail.com  Fri Jan 13 20:25:30 2012
From: alabdelouahab at gmail.com (aliane abdelouahab)
Date: Fri, 13 Jan 2012 21:25:30 +0100
Subject: nginx for windows 64 (not the cygwin)
In-Reply-To: <201201130141.05751.ne@vbart.ru>
References: <dfa4c8912a3627d4b9526974c1f7c09b.NginxMailingListEnglish@forum.nginx.org>
	<201201121803.21728.ne@vbart.ru>
	<CAP4kiWqXh+u3xBEoTAyToJv=cixJPD5W4NMrNuMZOzYt5qorcQ@mail.gmail.com>
	<201201130141.05751.ne@vbart.ru>
Message-ID: <CAP4kiWo2xy9-ekJT+6NRG+OWZxv6VGYuPOMkxFRo+YBCEQ-Hag@mail.gmail.com>

thank you for the reply
as a final solution, i've installed opensuse and i'll try to use it only
when it is possible.....

2012/1/12 Valentin V. Bartenev <ne at vbart.ru>

> On Friday 13 January 2012 00:15:02 aliane abdelouahab wrote:
> > so the default module are not included?
> > extras for example like ssl
> >
>
> You may download and check it yourself by using the "-V" option.
>
> TLS SNI support enabled
> configure arguments:
> --with-cc=cl
> --builddir=objs.msvc8
> --with-debug
> --prefix=
> --conf-path=conf/nginx.conf
> --pid-path=logs/nginx.pid
> --http-log-path=logs/access.log
> --error-log-path=logs/error.log
> --sbin-path=nginx.exe
> --http-client-body-temp-path=temp/client_body_temp
> --http-proxy-temp-path=temp/proxy_temp
> --http-fastcgi-temp-path=temp/fastcgi_temp
> --with-cc-opt=-DFD_SETSIZE=1024
> --with-pcre=objs.msvc8/lib/pcre-8.12
> --with-zlib=objs.msvc8/lib/zlib-1.2.5
> --with-select_module
> --with-http_realip_module
> --with-http_addition_module
> --with-http_sub_module
> --with-http_dav_module
> --with-http_stub_status_module
> --with-http_flv_module
> --with-http_mp4_module
> --with-http_gzip_static_module
> --with-http_random_index_module
> --with-http_secure_link_module
> --with-mail
> --with-openssl=objs.msvc8/lib/openssl-1.0.0e
> --with-openssl-opt=enable-tlsext
> --with-http_ssl_module
> --with-mail_ssl_module
> --with-ipv6
>
> wbr, Valentin V. Bartenev
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/980c0299/attachment.html>

From varia at e-healthexpert.org  Fri Jan 13 20:48:01 2012
From: varia at e-healthexpert.org (Mark Alan)
Date: Fri, 13 Jan 2012 20:48:01 +0000
Subject: Simple Groupware nginx config
Message-ID: <20120113204801.32e45b04@e-healthexpert.org>

In case anyone needs to setup Simple Groupware
http://www.simple-groupware.de/  (it is a web-based groupware that
offers email, calendaring, contacts, tasks, document management,
project management, synchronization with Outlook and cell phones,
full-text search, extensions, etc),
using Debian 6 / Ubuntu 11.10 and Nginx 1.1.12,

here is the config file provided by Simple Groupware's leader & main
developer:

(ED: it works, but it must be one of the ugliest Nginx config files that
we have seen)

# nginx - /etc/nginx/sites-available/sgw
server {
	listen 80;
	server_name sgw.localhost;
	#root sgw; # TODO change to Simple Groupware root directory
	access_log /var/log/nginx/sgw.access.log;
	error_log /var/log/nginx/sgw.error.log;
	root /var/www/sgw;
	#index index.php;
	
	# WebDAV server
	location ~ ^/sgdav {
	rewrite . /bin/webdav.php;
	}
	# CMS real URLs
	location ~ ^/cms/ {
		rewrite ^/cms/ext/(.*)$ /bin/ext/cms/$1 last;
		rewrite ^/cms/thumbs/(.*)$ /bin/preview.php?filename=$1
last; rewrite ^/cms/(.*?)/file/(.*)$ /bin/cms.php?page=$1&file=$2 last;
		rewrite ^/cms/(.*)$ /bin/cms.php?page=$1 last; 
	}
	# Root
	location = / {
		if ($request_method = "OPTIONS") {
			rewrite . /bin/webdav.php last;
		}
		try_files /bin/index.php /src/index.php /sgw_installer.php
=404; include /etc/nginx/fastcgi_params;
		fastcgi_pass unix:/tmp/php-fpm.socket;
		#fastcgi_index index.php;
		fastcgi_param SCRIPT_FILENAME
$document_root$fastcgi_script_name; }
	# Root PHP /*.php
	location ~ ^/([^/]+\.php)$ {
		try_files /bin/$1 /src/$1 $uri =404;
		#include fastcgi_params;
		include /etc/nginx/fastcgi_params;
		fastcgi_pass unix:/tmp/php-fpm.socket;
		#fastcgi_index index.php;
		fastcgi_param SCRIPT_FILENAME
$document_root$fastcgi_script_name; #fastcgi_pass   127.0.0.1:9000; #
use spawn-fcgi (!!) }
	# sgw src/*.php bin/*.php
	location ~ ^/(src|bin)/([^/]+\.php|ext/.+\.php)$ {
		include fastcgi_params;
		include /etc/nginx/fastcgi_params;
		fastcgi_pass unix:/tmp/php-fpm.socket;
		#fastcgi_index index.php;
		fastcgi_param SCRIPT_FILENAME
$document_root$fastcgi_script_name; #fastcgi_pass   127.0.0.1:9000; #
use spawn-fcgi (!!) }
	# Redirect static files
	location ~ ^/(src/|bin/)?(ext/.*|docs/.*)$ {
		try_files /custom/$2 /ext/$2 /bin/$2 /src/$2 $uri =404;
	}
	# Drop all other stuff
	location / {
		if (!-f $request_filename) { return 404; }
			return 403;
		}
}
# --- ends here ---

M.


From varia at e-healthexpert.org  Fri Jan 13 21:03:22 2012
From: varia at e-healthexpert.org (Mark Alan)
Date: Fri, 13 Jan 2012 21:03:22 +0000
Subject: ownCloud nginx config
Message-ID: <20120113210322.25d5f06a@e-healthexpert.org>

In case anyone needs to setup ownCloud
http://owncloud.org/  (it is a web-based personal cloud),
using Debian 6 / Ubuntu 11.10 and Nginx 1.1.12,

here is the config file provided by the ownCloud team:

# nginx - /etc/nginx/sites-available/owncloud
# nginx must be compiled using --with-http_dav_module,
server {
	listen 80;
	keepalive_timeout 70;
	server_name cloud.localhost;
	root /var/www/owncloud;
	client_max_body_size 8M;
	#
	dav_methods PUT DELETE MKCOL COPY MOVE;
	create_full_put_path on;
	dav_access user:rw group:rw all:r;
	#
	location / {
		index index.html index.htm index.php;
	}
	location ~ \.php$ {
		include /etc/nginx/fastcgi_params;
		fastcgi_pass unix:/tmp/php-fpm.socket;
		fastcgi_index index.php;
		fastcgi_param SCRIPT_FILENAME
$document_root$fastcgi_script_name; }
	location /owncloud {
		index  index.php;
		try_files $uri $uri/ @webdav;
	}
	location @webdav {
		fastcgi_split_path_info ^(.+\.php)(/.+)$;
		fastcgi_param SCRIPT_FILENAME
$document_root$fastcgi_script_name; include fastcgi_params;
		fastcgi_pass unix:/tmp/php-fpm.socket;
	}
}
# --- ends here ---

M.


From nginx-forum at nginx.us  Fri Jan 13 21:52:05 2012
From: nginx-forum at nginx.us (voidandany)
Date: Fri, 13 Jan 2012 16:52:05 -0500
Subject: Protect a specific php file
In-Reply-To: <20120113182934.GA27443@craic.sysops.org>
References: <20120113182934.GA27443@craic.sysops.org>
Message-ID: <7c2b7037e23bdd46d1dd256864fbb501.NginxMailingListEnglish@forum.nginx.org>

It seems resolved, during my test I have downloaded the php file and my
browser put it in cache so every time the web server respond the browser
serve the file in cache.
It explain why it worked with IE (file not in cache) but not with
Chrome...

Sorry for losing your time and many thanks to you Francis for the time
you spent making test

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221094,221213#msg-221213


From cloos at jhcloos.com  Fri Jan 13 21:54:38 2012
From: cloos at jhcloos.com (James Cloos)
Date: Fri, 13 Jan 2012 16:54:38 -0500
Subject: NGINX CSS optimizer module?
In-Reply-To: <CAPAQPqd2O0Ldd91P6gc69zf7aXjiES6TNzh3mfRfsUJteWOBAA@mail.gmail.com>
	(Casey Jordan's message of "Fri, 13 Jan 2012 13:10:45 -0500")
References: <CAPAQPqdzSUmvnEFPn3bbaCCE11WMNF2gDwRxpgwA30mQfSJgkg@mail.gmail.com>
	<201201132208.09158.ne@vbart.ru>
	<CAPAQPqd2O0Ldd91P6gc69zf7aXjiES6TNzh3mfRfsUJteWOBAA@mail.gmail.com>
Message-ID: <m3mx9rnu89.fsf@jhcloos.com>

>>>>> "CJ" == Casey Jordan <casey.jordan at jorsek.com> writes:

CJ> Ideally, on production server it would cache the result and serve
CJ> that on all following requests.

CJ> However, on development servers we would need to disable caching of
CJ> the files so that we could see our changes take effect in real time.

A better alternative may be to manage them just like any software
package.  Write a Makefile so that >make< links the various src css
files into a single file, and >make install< installs that into the
served directory.

-JimC
-- 
James Cloos <cloos at jhcloos.com>         OpenPGP: 1024D/ED7DAEA6


From mdounin at mdounin.ru  Fri Jan 13 22:20:25 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sat, 14 Jan 2012 02:20:25 +0400
Subject: proxy_intercept_errors and logging
In-Reply-To: <3a36efae80ac4fa8adf082759636ac86.NginxMailingListEnglish@forum.nginx.org>
References: <3a36efae80ac4fa8adf082759636ac86.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120113222025.GN67687@mdounin.ru>

Hello!

On Fri, Jan 13, 2012 at 02:47:00PM -0500, mikeyk wrote:

> Hi there,
> 
> We have nginx proxying to a bunch of django upstreams.
> 
> When those upstreams throw an error, we use proxy_intercept_errors to
> show a custom page. The problem is that this causes those 500s to be
> logged into the global access_log, rather than the access_log specified
> for that location. Here is a (reduced) example:
> 
>     # at the server 
>     access_log /logs/nginx/access.log accesslogformat;
>     proxy_intercept_errors on;
>     location ~ ^/oauth.*$ {
>         proxy_pass        http://appservers;
> 
>         access_log /logs/nginx/app.log applogformat;
>     }
> 
> Is there a way to get errors to log to the app.log in this case, even
> with proxy_intercept_errors to on? Right now we get all the 500s in the
> access.log.

Requests are logged in context of a location where their 
processing ends.  This means that after error_page redirection 
they will logged in the location where error_pages are processed.

If you want them to be logged to the same app.log, you need to 
define location for errors with the same logging.  To avoid mixing 
logs with normal errors, you may want to define separate 
error_page's for your apps.  Something like this should work: 

     location /oauth {
         proxy_pass ...

         proxy_intercept_errors on;
         error_page 502 504 /errors_oauth/5xx.html

         access_log /path/to/app.log applogformat;
     }

     location /errors_oauth/ {
         alias /path/to/errors/;
         access_log /path/to/app.log applogformat;
     }

Maxim Dounin


From casey.jordan at jorsek.com  Fri Jan 13 22:20:34 2012
From: casey.jordan at jorsek.com (Casey Jordan)
Date: Fri, 13 Jan 2012 17:20:34 -0500
Subject: NGINX CSS optimizer module?
In-Reply-To: <m3mx9rnu89.fsf@jhcloos.com>
References: <CAPAQPqdzSUmvnEFPn3bbaCCE11WMNF2gDwRxpgwA30mQfSJgkg@mail.gmail.com>
	<201201132208.09158.ne@vbart.ru>
	<CAPAQPqd2O0Ldd91P6gc69zf7aXjiES6TNzh3mfRfsUJteWOBAA@mail.gmail.com>
	<m3mx9rnu89.fsf@jhcloos.com>
Message-ID: <CAPAQPqfK8PkTDbzqGQbsWK8TCghyL_Kx2K_-Zik7yWZ5Jj3=5g@mail.gmail.com>

James,

Thanks for the suggestion and we do already do this for a good deal of our
application css.

The CSS I am talking about though is more dynamic, and is actually part of
publishing pipelines that users of our software configure and use when the
server is running. Users will be uploading new CSS, making changes etc, all
in real time to influence published outputs of their content.

So it's not ideal to compile this on deployment, which is a main reason I
need a more dynamic solution.

Thanks,

Casey

On Fri, Jan 13, 2012 at 4:54 PM, James Cloos <cloos at jhcloos.com> wrote:

> >>>>> "CJ" == Casey Jordan <casey.jordan at jorsek.com> writes:
>
> CJ> Ideally, on production server it would cache the result and serve
> CJ> that on all following requests.
>
> CJ> However, on development servers we would need to disable caching of
> CJ> the files so that we could see our changes take effect in real time.
>
> A better alternative may be to manage them just like any software
> package.  Write a Makefile so that >make< links the various src css
> files into a single file, and >make install< installs that into the
> served directory.
>
> -JimC
> --
> James Cloos <cloos at jhcloos.com>         OpenPGP: 1024D/ED7DAEA6
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>



-- 
--
Casey Jordan
easyDITA a product of Jorsek LLC
"CaseyDJordan" on LinkedIn, Twitter & Facebook
(585) 348 7399
easydita.com


This message is intended only for the use of the Addressee(s) and may
contain information that is privileged, confidential, and/or exempt from
disclosure under applicable law.  If you are not the intended recipient,
please be advised that any disclosure  copying, distribution, or use of
the information contained herein is prohibited.  If you have received
this communication in error, please destroy all copies of the message,
whether in electronic or hard copy format, as well as attachments, and
immediately contact the sender by replying to this e-mail or by phone.
Thank you.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/4d233fc1/attachment.html>

From zzz at zzz.org.ua  Fri Jan 13 22:45:54 2012
From: zzz at zzz.org.ua (Alexandr Gomoliako)
Date: Sat, 14 Jan 2012 00:45:54 +0200
Subject: NGINX CSS optimizer module?
In-Reply-To: <CAPAQPqfK8PkTDbzqGQbsWK8TCghyL_Kx2K_-Zik7yWZ5Jj3=5g@mail.gmail.com>
References: <CAPAQPqdzSUmvnEFPn3bbaCCE11WMNF2gDwRxpgwA30mQfSJgkg@mail.gmail.com>
	<201201132208.09158.ne@vbart.ru>
	<CAPAQPqd2O0Ldd91P6gc69zf7aXjiES6TNzh3mfRfsUJteWOBAA@mail.gmail.com>
	<m3mx9rnu89.fsf@jhcloos.com>
	<CAPAQPqfK8PkTDbzqGQbsWK8TCghyL_Kx2K_-Zik7yWZ5Jj3=5g@mail.gmail.com>
Message-ID: <CAKb9G9hS5bvhEJxddLvdN-NYYCWV11=i0VvGCynWb28P9GkR9A@mail.gmail.com>

On Sat, Jan 14, 2012 at 12:20 AM, Casey Jordan <casey.jordan at jorsek.com> wrote:

> I am looking for a feature which I hope NGINX supports, and that is
> optimization of serving CSS files, specifically around minimizing and
> resolving @import statements server side.
>
> For instance, to better manage the hundreds of css files we have, we
> started breaking them up and using the @import directive. Obviously this
> slows things down quite a bit because the browser has to make lots of
> additional requests.

> Ideally, on production server it would cache the result and serve
> that on all following requests.

Pretty trivial task for nginx-perl. Shouldn't take long to implement if you are
already familiar with embedded perl.


From zzz at zzz.org.ua  Fri Jan 13 22:50:15 2012
From: zzz at zzz.org.ua (Alexandr Gomoliako)
Date: Sat, 14 Jan 2012 00:50:15 +0200
Subject: NGINX CSS optimizer module?
In-Reply-To: <CAKb9G9hS5bvhEJxddLvdN-NYYCWV11=i0VvGCynWb28P9GkR9A@mail.gmail.com>
References: <CAPAQPqdzSUmvnEFPn3bbaCCE11WMNF2gDwRxpgwA30mQfSJgkg@mail.gmail.com>
	<201201132208.09158.ne@vbart.ru>
	<CAPAQPqd2O0Ldd91P6gc69zf7aXjiES6TNzh3mfRfsUJteWOBAA@mail.gmail.com>
	<m3mx9rnu89.fsf@jhcloos.com>
	<CAPAQPqfK8PkTDbzqGQbsWK8TCghyL_Kx2K_-Zik7yWZ5Jj3=5g@mail.gmail.com>
	<CAKb9G9hS5bvhEJxddLvdN-NYYCWV11=i0VvGCynWb28P9GkR9A@mail.gmail.com>
Message-ID: <CAKb9G9i8j1ZaUkiGQzzKVt7K3=V=jwsq1c1Nnxi8mftg2dhE-Q@mail.gmail.com>

> Pretty trivial task for nginx-perl. Shouldn't take long to implement if you are
> already familiar with embedded perl.

P.S. I think even original embedded perl would be enough, depending where
you store users' files.


From casey.jordan at jorsek.com  Fri Jan 13 22:58:53 2012
From: casey.jordan at jorsek.com (Casey Jordan)
Date: Fri, 13 Jan 2012 17:58:53 -0500
Subject: NGINX CSS optimizer module?
In-Reply-To: <CAKb9G9i8j1ZaUkiGQzzKVt7K3=V=jwsq1c1Nnxi8mftg2dhE-Q@mail.gmail.com>
References: <CAPAQPqdzSUmvnEFPn3bbaCCE11WMNF2gDwRxpgwA30mQfSJgkg@mail.gmail.com>
	<201201132208.09158.ne@vbart.ru>
	<CAPAQPqd2O0Ldd91P6gc69zf7aXjiES6TNzh3mfRfsUJteWOBAA@mail.gmail.com>
	<m3mx9rnu89.fsf@jhcloos.com>
	<CAPAQPqfK8PkTDbzqGQbsWK8TCghyL_Kx2K_-Zik7yWZ5Jj3=5g@mail.gmail.com>
	<CAKb9G9hS5bvhEJxddLvdN-NYYCWV11=i0VvGCynWb28P9GkR9A@mail.gmail.com>
	<CAKb9G9i8j1ZaUkiGQzzKVt7K3=V=jwsq1c1Nnxi8mftg2dhE-Q@mail.gmail.com>
Message-ID: <CAPAQPqeVUgXyvY7MXgTVjjnAqxM6-z0NnyZddKg9B2DUVTw2zw@mail.gmail.com>

Interesting. I have used perl before but not embedded perl.

I should also mention that NGINX is being used as a proxy server. So it
would not actually be reading files off the local drive.

Does this effect the use of embedded perl? Any examples you might be able
to point me to?

Thanks,

Casey

On Fri, Jan 13, 2012 at 5:50 PM, Alexandr Gomoliako <zzz at zzz.org.ua> wrote:

> > Pretty trivial task for nginx-perl. Shouldn't take long to implement if
> you are
> > already familiar with embedded perl.
>
> P.S. I think even original embedded perl would be enough, depending where
> you store users' files.
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>



-- 
--
Casey Jordan
easyDITA a product of Jorsek LLC
"CaseyDJordan" on LinkedIn, Twitter & Facebook
(585) 348 7399
easydita.com


This message is intended only for the use of the Addressee(s) and may
contain information that is privileged, confidential, and/or exempt from
disclosure under applicable law.  If you are not the intended recipient,
please be advised that any disclosure  copying, distribution, or use of
the information contained herein is prohibited.  If you have received
this communication in error, please destroy all copies of the message,
whether in electronic or hard copy format, as well as attachments, and
immediately contact the sender by replying to this e-mail or by phone.
Thank you.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120113/1f89fb94/attachment.html>

From nginx-forum at nginx.us  Fri Jan 13 23:27:46 2012
From: nginx-forum at nginx.us (mikeyk)
Date: Fri, 13 Jan 2012 18:27:46 -0500
Subject: proxy_intercept_errors and logging
In-Reply-To: <20120113222025.GN67687@mdounin.ru>
References: <20120113222025.GN67687@mdounin.ru>
Message-ID: <eb5dc21e5bc228b30a5377db1a864d7c.NginxMailingListEnglish@forum.nginx.org>

Thanks so much for the response--this was exactly what we were looking
for!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221201,221220#msg-221220


From zzz at zzz.org.ua  Fri Jan 13 23:41:41 2012
From: zzz at zzz.org.ua (Alexandr Gomoliako)
Date: Sat, 14 Jan 2012 01:41:41 +0200
Subject: NGINX CSS optimizer module?
In-Reply-To: <CAPAQPqeVUgXyvY7MXgTVjjnAqxM6-z0NnyZddKg9B2DUVTw2zw@mail.gmail.com>
References: <CAPAQPqdzSUmvnEFPn3bbaCCE11WMNF2gDwRxpgwA30mQfSJgkg@mail.gmail.com>
	<201201132208.09158.ne@vbart.ru>
	<CAPAQPqd2O0Ldd91P6gc69zf7aXjiES6TNzh3mfRfsUJteWOBAA@mail.gmail.com>
	<m3mx9rnu89.fsf@jhcloos.com>
	<CAPAQPqfK8PkTDbzqGQbsWK8TCghyL_Kx2K_-Zik7yWZ5Jj3=5g@mail.gmail.com>
	<CAKb9G9hS5bvhEJxddLvdN-NYYCWV11=i0VvGCynWb28P9GkR9A@mail.gmail.com>
	<CAKb9G9i8j1ZaUkiGQzzKVt7K3=V=jwsq1c1Nnxi8mftg2dhE-Q@mail.gmail.com>
	<CAPAQPqeVUgXyvY7MXgTVjjnAqxM6-z0NnyZddKg9B2DUVTw2zw@mail.gmail.com>
Message-ID: <CAKb9G9jKpEUUcWw8Nhuy+dDXEXC3nb3sitX1+1ya0u_u1afk-w@mail.gmail.com>

On Sat, Jan 14, 2012 at 12:58 AM, Casey Jordan <casey.jordan at jorsek.com> wrote:

> I should also mention that NGINX is being used as a proxy server.

Right, that was my first guess.

> Does this effect the use of embedded perl?

It does. You can't make requests to remote hosts asynchronously from original
embedded perl, but you can from nginx-perl.

> Any examples you might be able to point me to?

Documentation is the only place with simple examples right now:
http://zzzcpan.github.com/nginx-perl/Nginx.html#HTTP_API

I understand that you don't feel like using 3rd party distribution.
But it won't
be a problem, I'm going to generate an addon out of it in the future.


From nginx-forum at nginx.us  Sat Jan 14 00:13:27 2012
From: nginx-forum at nginx.us (locojohn)
Date: Fri, 13 Jan 2012 19:13:27 -0500
Subject: fastcgi_keep_conn consuming ports?
In-Reply-To: <FE4F5CE0-097D-4423-B77A-44F57DCD9CBC@nginx.com>
References: <FE4F5CE0-097D-4423-B77A-44F57DCD9CBC@nginx.com>
Message-ID: <f06772ad185c60f30866d37e6791b89b.NginxMailingListEnglish@forum.nginx.org>

Thank you Andrew!

Andrejs

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220911,221222#msg-221222


From cliff at develix.com  Sat Jan 14 01:34:52 2012
From: cliff at develix.com (Cliff Wells)
Date: Fri, 13 Jan 2012 17:34:52 -0800
Subject: ownCloud nginx config
In-Reply-To: <20120113210322.25d5f06a@e-healthexpert.org>
References: <20120113210322.25d5f06a@e-healthexpert.org>
Message-ID: <1326504892.2341.1.camel@portable-evil>

It would be more helpful if you documented these on the wiki:

http://wiki.nginx.org/Configuration#Applications


Regards,
Clif


On Fri, 2012-01-13 at 21:03 +0000, Mark Alan wrote:
> In case anyone needs to setup ownCloud
> http://owncloud.org/  (it is a web-based personal cloud),
> using Debian 6 / Ubuntu 11.10 and Nginx 1.1.12,
> 
> here is the config file provided by the ownCloud team:
> 
> # nginx - /etc/nginx/sites-available/owncloud
> # nginx must be compiled using --with-http_dav_module,
> server {
> 	listen 80;
> 	keepalive_timeout 70;
> 	server_name cloud.localhost;
> 	root /var/www/owncloud;
> 	client_max_body_size 8M;
> 	#
> 	dav_methods PUT DELETE MKCOL COPY MOVE;
> 	create_full_put_path on;
> 	dav_access user:rw group:rw all:r;
> 	#
> 	location / {
> 		index index.html index.htm index.php;
> 	}
> 	location ~ \.php$ {
> 		include /etc/nginx/fastcgi_params;
> 		fastcgi_pass unix:/tmp/php-fpm.socket;
> 		fastcgi_index index.php;
> 		fastcgi_param SCRIPT_FILENAME
> $document_root$fastcgi_script_name; }
> 	location /owncloud {
> 		index  index.php;
> 		try_files $uri $uri/ @webdav;
> 	}
> 	location @webdav {
> 		fastcgi_split_path_info ^(.+\.php)(/.+)$;
> 		fastcgi_param SCRIPT_FILENAME
> $document_root$fastcgi_script_name; include fastcgi_params;
> 		fastcgi_pass unix:/tmp/php-fpm.socket;
> 	}
> }
> # --- ends here ---
> 
> M.
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx



From nginx-forum at nginx.us  Sat Jan 14 13:27:53 2012
From: nginx-forum at nginx.us (double)
Date: Sat, 14 Jan 2012 08:27:53 -0500 (EST)
Subject: Limit IPv4/24 connections
Message-ID: <59320cebf989dafbfe05e93ab1abdd6b.NginxMailingListEnglish@forum.nginx.org>

Hello,

I would like to limit the connections to 100 per 255.255.255.0
Is there a chance e.g. to zero the last byte of "$binary_remote_addr"?

Something like:
set $remote_addr_24 "/$remote_addr/\.[0-9]*$/.0/"
limit_zone connectiontrack $remote_addr_24  16m;
limit_conn connectiontrack 100;

Thanks a lot
Markus

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221225,221225#msg-221225


From nginx-forum at nginx.us  Sat Jan 14 13:53:00 2012
From: nginx-forum at nginx.us (CoreTex)
Date: Sat, 14 Jan 2012 08:53:00 -0500
Subject: nginx 99% cpu if upstream server address is ipv6
In-Reply-To: <BANLkTimQUQvVcL++HyOMciTojtqS9xbNfg@mail.gmail.com>
References: <BANLkTimQUQvVcL++HyOMciTojtqS9xbNfg@mail.gmail.com>
Message-ID: <7aaf776f7669b52d1c182fd1dbaae82d.NginxMailingListEnglish@forum.nginx.org>

Hi,
i have the same problem. any new solutions ?

Regards, Eric

My Setup is:

WORLD -> ( nginx  ipv4 blancer ) -> ( 3 ipv4 webserver )
WORLD -> ( nginx  ipv6 blancer ) -> ( 3 ipv6 webserver )

My Config is:
upstream lb4 {
    server XXX.XXX.XXX.XXX:80;
    server XXX.XXX.XXX.XXX:80;
    server XXX.XXX.XXX.XXX:80;
}
upstream lb6 {
    server [xxxx:xxxx:xxxx::xxxx]:80;
    server [xxxx:xxxx:xxxx::xxxx]:80;
    server [xxxx:xxxx:xxxx::xxxx]:80;
}
server {
    listen  80; ## listen for ipv4
    server_name lb.netsyn.eu;

    access_log  /var/log/nginx/xxx.xxx.eu.access.log;

    location / {
            proxy_pass      http://lb4;
            proxy_redirect  off;
            proxy_next_upstream     error timeout invalid_header
http_500;
            proxy_connect_timeout   2;
            proxy_set_header        Host            $host;
            proxy_set_header        X-Real-IP       $remote_addr;
            proxy_set_header        X-Forwarded-For
$proxy_add_x_forwarded_for;
    }
}

server {
    listen   [::]:80 default ipv6only=on; ## listen for ipv6
    server_name lb.netsyn.eu;

    access_log  /var/log/nginx/xxx.xxxx.eu.access.log;

    location / {
            proxy_pass      http://lb6;
            proxy_redirect  off;
            proxy_next_upstream     error timeout invalid_header
http_500;
            proxy_connect_timeout   2;
            proxy_set_header        Host            $host;
            proxy_set_header        X-Real-IP       $remote_addr;
            proxy_set_header        X-Forwarded-For
$proxy_add_x_forwarded_for;
    }
}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,208399,221226#msg-221226


From ne at vbart.ru  Sat Jan 14 14:04:00 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Sat, 14 Jan 2012 18:04:00 +0400
Subject: Limit IPv4/24 connections
In-Reply-To: <59320cebf989dafbfe05e93ab1abdd6b.NginxMailingListEnglish@forum.nginx.org>
References: <59320cebf989dafbfe05e93ab1abdd6b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <201201141804.00602.ne@vbart.ru>

On Saturday 14 January 2012 17:27:53 double wrote:
> Hello,
> 
> I would like to limit the connections to 100 per 255.255.255.0
> Is there a chance e.g. to zero the last byte of "$binary_remote_addr"?
> 
> Something like:
> set $remote_addr_24 "/$remote_addr/\.[0-9]*$/.0/"
> limit_zone connectiontrack $remote_addr_24  16m;
> limit_conn connectiontrack 100;
> 

   # this captures only first 3 bytes of $binary_remote_addr:

   if ($binary_remote_addr ~ "^(\C{3})") {
       set $remote_addr_24 $1;
   }


 wbr, Valentin V. Bartenev


From ne at vbart.ru  Sat Jan 14 14:11:34 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Sat, 14 Jan 2012 18:11:34 +0400
Subject: Limit IPv4/24 connections
In-Reply-To: <201201141804.00602.ne@vbart.ru>
References: <59320cebf989dafbfe05e93ab1abdd6b.NginxMailingListEnglish@forum.nginx.org>
	<201201141804.00602.ne@vbart.ru>
Message-ID: <201201141811.34685.ne@vbart.ru>

On Saturday 14 January 2012 18:04:00 Valentin V. Bartenev wrote:
> On Saturday 14 January 2012 17:27:53 double wrote:
> > Hello,
> > 
> > I would like to limit the connections to 100 per 255.255.255.0
> > Is there a chance e.g. to zero the last byte of "$binary_remote_addr"?
> > 
> > Something like:
> > set $remote_addr_24 "/$remote_addr/\.[0-9]*$/.0/"
> > limit_zone connectiontrack $remote_addr_24  16m;
> > limit_conn connectiontrack 100;
> 
>    # this captures only first 3 bytes of $binary_remote_addr:
> 
>    if ($binary_remote_addr ~ "^(\C{3})") {
>        set $remote_addr_24 $1;
>    }
> 
> 

 or:

 map $binary_remote_addr $remote_addr_24 {
     default             "";
     ~^(?P<bytes>\C{3})  $bytes;
 }

 wbr, Valentin V. Bartenev


From jelledejong at powercraft.nl  Sat Jan 14 15:17:37 2012
From: jelledejong at powercraft.nl (Jelle de Jong)
Date: Sat, 14 Jan 2012 16:17:37 +0100
Subject: multiple https proxy_pass to diffrent apache server
Message-ID: <4F119C91.6070208@powercraft.nl>

Hello everybody,

The relevant part of the config: http://paste.debian.net/hidden/9d3d8f16/

I want to proxy pass port 443 based on the server_name to different
Apache servers, anybody knows how should I do this?

Summary of the config: (see the above pastebin for more info)

server {
    listen      443;
    server_name calendar.powercraft.nl;
    ssl         on;
    ssl_certificate     /etc/ssl/certs/certificate.pem;
    ssl_certificate_key /etc/ssl/certs/privatekey.pem;
    keepalive_timeout   70;
    location / {
        proxy_redirect      off;
        proxy_pass          https://192.168.24.66/;
    }
}

server {
    listen      443;
    server_name webmail.tuxblogger.nl;
    ssl         on;
    ssl_certificate     /etc/ssl/certs/certificate.pem;
    ssl_certificate_key /etc/ssl/certs/privatekey.pem;
    location / {
        proxy_redirect      off;
        proxy_pass          https://192.168.24.67/;
    }
}

I want to add several more server_names if possible.

Thanks in advance,

Kind regards,

Jelle


From ne at vbart.ru  Sat Jan 14 16:33:45 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Sat, 14 Jan 2012 20:33:45 +0400
Subject: multiple https proxy_pass to diffrent apache server
In-Reply-To: <4F119C91.6070208@powercraft.nl>
References: <4F119C91.6070208@powercraft.nl>
Message-ID: <201201142033.45830.ne@vbart.ru>

On Saturday 14 January 2012 19:17:37 Jelle de Jong wrote:
> Hello everybody,
> 
> The relevant part of the config: http://paste.debian.net/hidden/9d3d8f16/
> 
> I want to proxy pass port 443 based on the server_name to different
> Apache servers, anybody knows how should I do this?
> 
> Summary of the config: (see the above pastebin for more info)
> 
> server {
>     listen      443;
>     server_name calendar.powercraft.nl;
>     ssl         on;
>     ssl_certificate     /etc/ssl/certs/certificate.pem;
>     ssl_certificate_key /etc/ssl/certs/privatekey.pem;
>     keepalive_timeout   70;
>     location / {
>         proxy_redirect      off;
>         proxy_pass          https://192.168.24.66/;
>     }
> }
> 
> server {
>     listen      443;
>     server_name webmail.tuxblogger.nl;
>     ssl         on;
>     ssl_certificate     /etc/ssl/certs/certificate.pem;
>     ssl_certificate_key /etc/ssl/certs/privatekey.pem;
>     location / {
>         proxy_redirect      off;
>         proxy_pass          https://192.168.24.67/;
>     }
> }
> 
> I want to add several more server_names if possible.
> 

Probably something like this would work:

 map $host $backend {
     calendar.powercraft.nl  192.168.24.66;
     webmail.tuxblogger.nl   192.168.24.67;
 }

 server {
    listen      443;
    server_name calendar.powercraft.nl webmail.tuxblogger.nl;

    ssl on;
    ssl_certificate     /etc/ssl/certs/certificate.pem;
    ssl_certificate_key /etc/ssl/certs/privatekey.pem;

    keepalive_timeout   70;

    location / {
        proxy_pass  https://$backend/;
    }
 }

 
 wbr, Valentin V. Bartenev



From ne at vbart.ru  Sat Jan 14 16:41:21 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Sat, 14 Jan 2012 20:41:21 +0400
Subject: multiple https proxy_pass to diffrent apache server
In-Reply-To: <201201142033.45830.ne@vbart.ru>
References: <4F119C91.6070208@powercraft.nl> <201201142033.45830.ne@vbart.ru>
Message-ID: <201201142041.21800.ne@vbart.ru>

On Saturday 14 January 2012 20:33:45 Valentin V. Bartenev wrote:
[...]
> 
>     location / {
>         proxy_pass  https://$backend/;
>     }

 Oops, correct:
 
 proxy_pass  https://$backend$uri;


 wbr, Valentin V. Bartenev


From gmm at csdoc.com  Sat Jan 14 16:48:23 2012
From: gmm at csdoc.com (Gena Makhomed)
Date: Sat, 14 Jan 2012 18:48:23 +0200
Subject: multiple https proxy_pass to diffrent apache server
In-Reply-To: <4F119C91.6070208@powercraft.nl>
References: <4F119C91.6070208@powercraft.nl>
Message-ID: <4F11B1D7.4020309@csdoc.com>

On 14.01.2012 17:17, Jelle de Jong wrote:

> The relevant part of the config: http://paste.debian.net/hidden/9d3d8f16/

> I want to proxy pass port 443 based on the server_name to different
> Apache servers, anybody knows how should I do this?

http://nginx.org/en/docs/http/configuring_https_servers.html

> server {
>      listen      443;
>      server_name calendar.powercraft.nl;
>      ssl         on;
>      ssl_certificate     /etc/ssl/certs/certificate.pem;
>      ssl_certificate_key /etc/ssl/certs/privatekey.pem;
>      keepalive_timeout   70;
>      location / {
>          proxy_redirect      off;
>          proxy_pass          https://192.168.24.66/;
>      }
> }
>
> server {
>      listen      443;
>      server_name webmail.tuxblogger.nl;
>      ssl         on;
>      ssl_certificate     /etc/ssl/certs/certificate.pem;
>      ssl_certificate_key /etc/ssl/certs/privatekey.pem;
>      location / {
>          proxy_redirect      off;
>          proxy_pass          https://192.168.24.67/;
>      }
> }
>
> I want to add several more server_names if possible.

yes, it is possible.

make separate server{ ... } for each https site.

probably you need separate IP for each site,
or use only SNI-capable clients (browsers)

-- 
Best regards,
  Gena


From ne at vbart.ru  Sat Jan 14 17:58:03 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Sat, 14 Jan 2012 21:58:03 +0400
Subject: multiple https proxy_pass to diffrent apache server
In-Reply-To: <201201142041.21800.ne@vbart.ru>
References: <4F119C91.6070208@powercraft.nl> <201201142033.45830.ne@vbart.ru>
	<201201142041.21800.ne@vbart.ru>
Message-ID: <201201142158.03496.ne@vbart.ru>

On Saturday 14 January 2012 20:41:21 Valentin V. Bartenev wrote:
> On Saturday 14 January 2012 20:33:45 Valentin V. Bartenev wrote:
> [...]
> 
> >     location / {
> >     
> >         proxy_pass  https://$backend/;
> >     
> >     }
> 
>  Oops, correct:
> 
>  proxy_pass  https://$backend$uri;
> 

And, probably, you didn't want proxy pass HTTPS to upstream:

 proxy_pass  http://$backend$uri;

I also expected that you have multi-domain certificate.


 wbr, Valentin V. Bartenev


From lists at ruby-forum.com  Sat Jan 14 20:13:06 2012
From: lists at ruby-forum.com (xe2 delphi)
Date: Sat, 14 Jan 2012 21:13:06 +0100
Subject: needs nginx compile install for use 3rd party module?
Message-ID: <997a1821483313965209a5358ee8d930@ruby-forum.com>

I know I have to compile 3rd party module if I want to use it.

And I want to know if I have to compile and install Nginx to use 3rd
party module.

Please let me know.

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


From appa at perusio.net  Sat Jan 14 20:35:54 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Sat, 14 Jan 2012 20:35:54 +0000
Subject: needs nginx compile install for use 3rd party module?
In-Reply-To: <997a1821483313965209a5358ee8d930@ruby-forum.com>
References: <997a1821483313965209a5358ee8d930@ruby-forum.com>
Message-ID: <87zkdqvx79.wl%appa@perusio.net>

On 14 Jan 2012 20h13 WET, lists at ruby-forum.com wrote:

> I know I have to compile 3rd party module if I want to use it.
>
> And I want to know if I have to compile and install Nginx to use 3rd
> party module.
>
> Please let me know.

Usually the module has some form of documentation where
installation instructions are provided. 

Generically you need a config file and to pass the nginx configure
script a --add-module argument for the module you want to compile in:

       --add-module=/path/to/module

--- appa


From mdounin at mdounin.ru  Sat Jan 14 23:26:21 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sun, 15 Jan 2012 03:26:21 +0400
Subject: multiple https proxy_pass to diffrent apache server
In-Reply-To: <201201142041.21800.ne@vbart.ru>
References: <4F119C91.6070208@powercraft.nl> <201201142033.45830.ne@vbart.ru>
	<201201142041.21800.ne@vbart.ru>
Message-ID: <20120114232621.GQ67687@mdounin.ru>

Hello!

On Sat, Jan 14, 2012 at 08:41:21PM +0400, Valentin V. Bartenev wrote:

> On Saturday 14 January 2012 20:33:45 Valentin V. Bartenev wrote:
> [...]
> > 
> >     location / {
> >         proxy_pass  https://$backend/;
> >     }
> 
>  Oops, correct:
>  
>  proxy_pass  https://$backend$uri;

The $uri part isn't needed, just

   proxy_pass  https://$backend;

is enough.

Maxim Dounin


From harish at octopart.com  Sun Jan 15 02:48:36 2012
From: harish at octopart.com (Harish Agarwal)
Date: Sat, 14 Jan 2012 20:48:36 -0600
Subject: 301 Redirects Break Long Posts?
Message-ID: <CAGvzxGiRoVXdRMkLQmGYzOFKZO4QgrFC-7L3gygVvsosRF1hqg@mail.gmail.com>

Hello,

I'm running nginx 1.0.10 with a few name based virtual hosts configured.
 There are a few permanent redirects configured of the format:

    server {
        listen       80;
        server_name  domain.com;
        rewrite ^/(.*) http://www.domain.com/$1 permanent;
    }

The site sees frequent POSTs to it, which in turn seemingly randomly
disconnect from the upstream server after sending the data along (the
request completes successfully upstream and then complains of a broken pipe
as nginx has disconnected).  The disconnects seem highly coupled to 301
redirects - in the debug logs it appears that every POST is interrupted by
a permanent redirect (and that the thread serving the request stops serving
requests entirely after the redirect finishes, the last debug statement
being "http log handler").  After removing the server block the POSTs run
without any problems.  I'm wondering if this is enough information to
arouse anyone's suspicions as to what the problem is?  Any help would be
greatly appreciated.

-Harish
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120114/13473678/attachment.html>

From appa at perusio.net  Sun Jan 15 03:00:31 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Sun, 15 Jan 2012 03:00:31 +0000
Subject: 301 Redirects Break Long Posts?
In-Reply-To: <CAGvzxGiRoVXdRMkLQmGYzOFKZO4QgrFC-7L3gygVvsosRF1hqg@mail.gmail.com>
References: <CAGvzxGiRoVXdRMkLQmGYzOFKZO4QgrFC-7L3gygVvsosRF1hqg@mail.gmail.com>
Message-ID: <87wr8twtyo.wl%appa@perusio.net>

On 15 Jan 2012 02h48 WET, harish at octopart.com wrote:

> [1  <multipart/alternative (7bit)>]
> [1.1  <text/plain; ISO-8859-1 (7bit)>]
> Hello,
>
> I'm running nginx 1.0.10 with a few name based virtual hosts
> configured.  There are a few permanent redirects configured of the
> format:
>
> server {
> listen       80;
> server_name  domain.com;
> rewrite ^/(.*) http://www.domain.com/$1 permanent;
> }

To do that type of thing try this and see of it helps:

server {
    listen 80;
    server_name domain.com;
    return 301 http://www.domain.com$request_uri;
}

There's no need for any regex capture or rewrite for that matter. The
$request_uri variable has both the uri and the query string.

> The site sees frequent POSTs to it, which in turn seemingly randomly
> disconnect from the upstream server after sending the data along
> (the request completes successfully upstream and then complains of a
> broken pipe as nginx has disconnected).  The disconnects seem highly
> coupled to 301 redirects - in the debug logs it appears that every
> POST is interrupted by a permanent redirect (and that the thread
> serving the request stops serving requests entirely after the
> redirect finishes, the last debug statement being "http log
> handler").  After removing the server block the POSTs run without
> any problems.  I'm wondering if this is enough information to arouse
> anyone's suspicions as to what the problem is?  Any help would be
> greatly appreciated.

--- appa


From harish at octopart.com  Sun Jan 15 03:13:00 2012
From: harish at octopart.com (Harish Agarwal)
Date: Sat, 14 Jan 2012 21:13:00 -0600
Subject: 301 Redirects Break Long Posts?
In-Reply-To: <87wr8twtyo.wl%appa@perusio.net>
References: <CAGvzxGiRoVXdRMkLQmGYzOFKZO4QgrFC-7L3gygVvsosRF1hqg@mail.gmail.com>
	<87wr8twtyo.wl%appa@perusio.net>
Message-ID: <CAGvzxGhfrm=24wRdGrRwoiV0UySy3_tetZyiqLwi9OOm3XbwpA@mail.gmail.com>

Hi Antonio,

Thanks for the tip!  I was serving 301s in this way before and was still
experiencing the problem mentioned.  I'll revert the configuration for now
but any other advice would be great!

Thanks,
-Harish

On Sat, Jan 14, 2012 at 9:00 PM, Ant?nio P. P. Almeida <appa at perusio.net>wrote:

> On 15 Jan 2012 02h48 WET, harish at octopart.com wrote:
>
> > [1  <multipart/alternative (7bit)>]
> > [1.1  <text/plain; ISO-8859-1 (7bit)>]
> > Hello,
> >
> > I'm running nginx 1.0.10 with a few name based virtual hosts
> > configured.  There are a few permanent redirects configured of the
> > format:
> >
> > server {
> > listen       80;
> > server_name  domain.com;
> > rewrite ^/(.*) http://www.domain.com/$1 permanent;
> > }
>
> To do that type of thing try this and see of it helps:
>
> server {
>    listen 80;
>    server_name domain.com;
>     return 301 http://www.domain.com$request_uri;
> }
>
> There's no need for any regex capture or rewrite for that matter. The
> $request_uri variable has both the uri and the query string.
>
> > The site sees frequent POSTs to it, which in turn seemingly randomly
> > disconnect from the upstream server after sending the data along
> > (the request completes successfully upstream and then complains of a
> > broken pipe as nginx has disconnected).  The disconnects seem highly
> > coupled to 301 redirects - in the debug logs it appears that every
> > POST is interrupted by a permanent redirect (and that the thread
> > serving the request stops serving requests entirely after the
> > redirect finishes, the last debug statement being "http log
> > handler").  After removing the server block the POSTs run without
> > any problems.  I'm wondering if this is enough information to arouse
> > anyone's suspicions as to what the problem is?  Any help would be
> > greatly appreciated.
>
> --- appa
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120114/cc55e1f9/attachment.html>

From lists at ruby-forum.com  Sun Jan 15 04:45:20 2012
From: lists at ruby-forum.com (xe2 delphi)
Date: Sun, 15 Jan 2012 05:45:20 +0100
Subject: needs nginx compile install for use 3rd party module?
In-Reply-To: <997a1821483313965209a5358ee8d930@ruby-forum.com>
References: <997a1821483313965209a5358ee8d930@ruby-forum.com>
Message-ID: <02a8bff18d36727799310339891e799e@ruby-forum.com>

I understand that I need add a compile option!
thanks!!

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


From quintinpar at gmail.com  Sun Jan 15 06:57:41 2012
From: quintinpar at gmail.com (Quintin Par)
Date: Sun, 15 Jan 2012 12:27:41 +0530
Subject: Cache 404 & 500 pages
Message-ID: <CACW9FdNn-ckaqjLWWeT8GGPjoCgi4iAKq_sjKLSuxKiR024XTg@mail.gmail.com>

Hi all,

When I attempt caching these pages as shown below

        error_page  404        /404.html;

        location = /404.html {

                    root /var/www/prod /templates/;

                    internal;

                    add_header Cache-Control public;

                    add_header X-Cache-Status $upstream_cache_status;

                    proxy_cache             generalcache;

                    proxy_cache_key         backend$request_uri;

                    proxy_cache_use_stale   updating;

        }

I don?t see the X-Cache-Status header in the html response. Does that mean
the page does not get cached?

-Quintin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120115/98597dc1/attachment.html>

From mdounin at mdounin.ru  Sun Jan 15 11:12:41 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Sun, 15 Jan 2012 15:12:41 +0400
Subject: Cache 404 & 500 pages
In-Reply-To: <CACW9FdNn-ckaqjLWWeT8GGPjoCgi4iAKq_sjKLSuxKiR024XTg@mail.gmail.com>
References: <CACW9FdNn-ckaqjLWWeT8GGPjoCgi4iAKq_sjKLSuxKiR024XTg@mail.gmail.com>
Message-ID: <20120115111241.GS67687@mdounin.ru>

Hello!

On Sun, Jan 15, 2012 at 12:27:41PM +0530, Quintin Par wrote:

> Hi all,
> 
> When I attempt caching these pages as shown below
> 
>         error_page  404        /404.html;
> 
>         location = /404.html {
> 
>                     root /var/www/prod /templates/;
> 
>                     internal;
> 
>                     add_header Cache-Control public;
> 
>                     add_header X-Cache-Status $upstream_cache_status;
> 
>                     proxy_cache             generalcache;
> 
>                     proxy_cache_key         backend$request_uri;
> 
>                     proxy_cache_use_stale   updating;
> 
>         }
> 
> I don?t see the X-Cache-Status header in the html response. Does that mean
> the page does not get cached?

The "proxy_cache" directive works if you are using proxy_pass.  It 
doesn't work if you are serving static files as in the config 
above.  This is expected behaviour.

Maxim Dounin


From xb.zhou at qq.com  Sun Jan 15 17:03:58 2012
From: xb.zhou at qq.com (=?ISO-8859-1?B?WkhPVSBYaWFvYm8=?=)
Date: Mon, 16 Jan 2012 01:03:58 +0800
Subject: About nginx module developing
Message-ID: <tencent_319E5BF8669B9E703FC4C7E2@qq.com>

hi everyone:

I am new to nginx module developing.
I want to implement a module to deal with this demand:

1, nginx receives a request from client 
2, nginx sends two internal requests(http or protobuf etc.) to 2 backend servers
3, after the responses of the two internal requests received nginx makes the response and sends to the client

I am not sure that whether nginx can do this.
thx


------------------
Sincerely yours
                         ZHOU Xiaobo
<br>

From nginx-forum at nginx.us  Sun Jan 15 19:40:01 2012
From: nginx-forum at nginx.us (dbanks)
Date: Sun, 15 Jan 2012 14:40:01 -0500
Subject: Inconsistent latency to sub_status
Message-ID: <bd7c9b57b6cc36a0f64b87bc5736aef7.NginxMailingListEnglish@forum.nginx.org>

Hi,

We use nginx in a latency sensitive application on our own hardware.  We
have a monitoring system running on the local network that pings the
nginx_status page from the sub_status module once per minute.  In
addition to collecting the stats from the page, it also measures the
round trip latency.  Latency is often sub 1ms, but sometimes spikes to
over 30ms or more and can stay there for up to 5 consecutive minutes.  I
am trying to identify the source of the latency jitter, if possible.

We run ubuntu 10.04, which has a 100Hz system timer, so I expect that
consistently getting below 10ms response is unlikely.  Spiking to 30ms,
however, feels like there is a config issue somewhere - whether system
or nginx.  Nginx is a front end for us; all requests are passed to a 
FCGI backend via unix sockets (no static file serving).

This is an 8 physical core box with hyperthreading turned on.  It isn't
swapping, has very low disk activity, no TCP/IP issues that I can tell. 
CPU use ranges from 20 - 40% across all cores (the application and nginx
are co-resident).  The box handles between 300 and 1500 requests/sec.  I
have not noticed any correlation between higher overall load and higher
ping latency; it seems to be pretty random.

Any suggestions on other things to investigate are appreciated.

Config details follow.

Cheers,

Dean

We are running Nginx 1.0.5

nginx.conf (excerpt):
worker_processes  8;  
worker_rlimit_nofile 16384;   
worker_priority 0;

events {
        multi_accept off;  
        accept_mutex off;  
        worker_connections 8192;
}

http {
    error_log  /var/log/nginx/error.log info buffer=32k;
    access_log  /var/log/nginx/access.log main buffer=128k;
    sendfile        off;  
    tcp_nodelay on;  
}

server {
        listen  <IP>:80 default_server backlog=1024;
}



sysctl.conf (excerpt):

vm.swappiness=0
vm.vfs_cache_pressure = 50
fs.file-max = 65535
net.ipv4.ip_local_port_range = 2000 65000
net.ipv4.tcp_rmem = 4096 87380 8388608
net.ipv4.tcp_wmem = 4096 87380 8388608
net.core.rmem_max = 8388608
net.core.wmem_max = 8388608
net.core.netdev_max_backlog = 5000
net.core.somaxconn=4096
net.ipv4.tcp_max_syn_backlog = 2048
net.ipv4.tcp_window_scaling = 1
net.ipv4.tcp_fin_timeout = 30
net.unix.max_dgram_qlen=1024

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221270,221270#msg-221270


From mauro.stettler at gmail.com  Sun Jan 15 23:32:11 2012
From: mauro.stettler at gmail.com (Mauro Stettler)
Date: Mon, 16 Jan 2012 07:32:11 +0800
Subject: About nginx module developing
In-Reply-To: <tencent_319E5BF8669B9E703FC4C7E2@qq.com>
References: <tencent_319E5BF8669B9E703FC4C7E2@qq.com>
Message-ID: <CAP-8hjrbAPe_ZrMwQxdDzTf0OHy6oMDWdhJmkV1kAJoTd_3zoA@mail.gmail.com>

you can use the eval module to do this

On Mon, Jan 16, 2012 at 01:03, ZHOU Xiaobo <xb.zhou at qq.com> wrote:
> hi everyone:
>
> I am new to nginx module developing.
> I want to implement a module to deal with this demand:
>
> 1, nginx receives a request from client
> 2, nginx sends two internal requests(http or protobuf etc.) to 2 backend servers
> 3, after the responses of the two internal requests received nginx makes the response and sends to the client
>
> I am not sure that whether nginx can do this.
> thx
>
>
> ------------------
> Sincerely yours
> ? ? ? ? ? ? ? ? ? ? ? ? ZHOU Xiaobo
> <br>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From nginx-forum at nginx.us  Mon Jan 16 04:35:44 2012
From: nginx-forum at nginx.us (etrader)
Date: Sun, 15 Jan 2012 23:35:44 -0500
Subject: Regex rewrite to root document
Message-ID: <026ecd3865a95af1d3751ce80a2df5fd.NginxMailingListEnglish@forum.nginx.org>

I have a simple re-write of

rewrite ^/(.*) /post.php?id=$1 last;

This will breaks my main index page, as domain.com will also be
redirected to "post.php?id="

How can I modify my regular expression not to change the main index
page?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221278,221278#msg-221278


From appa at perusio.net  Mon Jan 16 04:59:03 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Mon, 16 Jan 2012 04:59:03 +0000
Subject: Regex rewrite to root document
In-Reply-To: <026ecd3865a95af1d3751ce80a2df5fd.NginxMailingListEnglish@forum.nginx.org>
References: <026ecd3865a95af1d3751ce80a2df5fd.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87r4z0w8dk.wl%appa@perusio.net>

On 16 Jan 2012 04h35 WET, nginx-forum at nginx.us wrote:

> I have a simple re-write of
>
> rewrite ^/(.*) /post.php?id=$1 last;
>
> This will breaks my main index page, as domain.com will also be
> redirected to "post.php?id="
>
> How can I modify my regular expression not to change the main index
> page?

Easier to define *another* location without any rewrite just for / :

location = / {
    ...
}

location / {
    # do the rewrite here
    ...
}

--- appa


From zhuzhaoyuan at gmail.com  Mon Jan 16 05:52:09 2012
From: zhuzhaoyuan at gmail.com (Joshua Zhu)
Date: Mon, 16 Jan 2012 13:52:09 +0800
Subject: [PATCH 1 of 2] fix sched_setaffinity(2) call failure
In-Reply-To: <20120113181651.GM67687@mdounin.ru>
References: <CAOKiq2D2hwCFy3+9wFeU8ZVxtHRgU3Nrj4-UafSuYhgn8AzCOw@mail.gmail.com>
	<20120113181651.GM67687@mdounin.ru>
Message-ID: <CAOKiq2AaxHNwB=+y1d=_x0vJE6YrZ3gC0=JGpwPPBF6OmN7bnQ@mail.gmail.com>

Hi,

On Sat, Jan 14, 2012 at 2:16 AM, Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
>
> On Thu, Jan 12, 2012 at 11:37:00PM +0800, Joshua Zhu wrote:
>
> > Hi,
> >
> > 'cpu_set_t' should be used instead of 'long' when calling
> > sched_setaffinity(2). Otherwise it may fail on some Linux systems.
> > Please see the attachment for more detail.
>
> According to sched_setaffinity(2) at least in kernels >= 2.6.9 it
> should be ok to use any length as long as bits above kernel mask
> used aren't referenced.
>
>
I've tested it on Linux 2.6.9 (Red Hat Enterprise Linux AS release 4) and
unfortunately sched_setaffinity(2) failed.


> I believe the real problem is that code uses 32 as cpusetsize,
> while this is the size in bytes.  Hence it references arbitrary
> memory and this causes EINVAL to be returned by glibc if there are
> bits referenced above kernel cpumask size, see
> sysdeps/unix/sysv/linux/sched_setaffinity.c in glibc's sources.
>
> The following patch should be enough:
>
> diff -r e1296af53cc0 src/os/unix/ngx_process_cycle.c
> --- a/src/os/unix/ngx_process_cycle.c   Mon Dec 26 00:00:00 2011 +0400
> +++ b/src/os/unix/ngx_process_cycle.c   Fri Jan 13 11:58:01 2012 -0500
> @@ -914,7 +914,10 @@
>         ngx_log_error(NGX_LOG_NOTICE, cycle->log, 0,
>                       "sched_setaffinity(0x%08Xl)", cpu_affinity);
>
> -        if (sched_setaffinity(0, 32, (cpu_set_t *) &cpu_affinity) == -1) {
> +        if (sched_setaffinity(0, sizeof(cpu_affinity),
> +                              (cpu_set_t *) &cpu_affinity)
> +            == -1)
> +        {
>             ngx_log_error(NGX_LOG_ALERT, cycle->log, ngx_errno,
>                           "sched_setaffinity(0x%08Xl) failed",
> cpu_affinity);
>         }
>
> Could you please test if it works for you?
>

Yes, it works on one of my Linux box which has two cores. But I have not
tested it on more other systems yet.


>
> It passes my limited testing, but I wasn't really able to
> reproduce the problem without modifying the sched_setaffinity()
> call in question to use bigger cpusetsize.
>
> [...]
>
> > diff -uprN nginx-1.1.12/src/core/nginx.c
> nginx-1.1.12-bugfix/src/core/nginx.c
> > --- nginx-1.1.12/src/core/nginx.c     2011-11-16 04:35:42.000000000 +0800
> > +++ nginx-1.1.12-bugfix/src/core/nginx.c      2012-01-12 11:06:00.000000000
> +0800
> > @@ -1245,15 +1245,15 @@ ngx_set_cpu_affinity(ngx_conf_t *cf, ngx
> >      ngx_core_conf_t  *ccf = conf;
> >
> >      u_char            ch;
> > -    u_long           *mask;
> > +    cpu_set_t        *mask;
> >      ngx_str_t        *value;
> > -    ngx_uint_t        i, n;
> > +    ngx_uint_t        i, j, n;
> >
> >      if (ccf->cpu_affinity) {
> >          return "is duplicate";
> >      }
> >
> > -    mask = ngx_palloc(cf->pool, (cf->args->nelts - 1) * sizeof(long));
> > +    mask = ngx_palloc(cf->pool, (cf->args->nelts - 1) *
> sizeof(cpu_set_t));
> >      if (mask == NULL) {
> >          return NGX_CONF_ERROR;
> >      }
>
> [...]
>
> While I generally agree that using cpu_set_t and CPU_* macros
> explicitly would be better, I don't think we want to introduce
> linux-centric code here in platform-independent code.
>
> Instead we probably want to leave this code generic enough to be
> useable with other scheduler interfaces, and convert the data
> somewhere near platform-specific whatever_setaffinity() calls if
> needed.  (Likely unneeded for the Linux with the above patch.)
>
>
We use cpu_set_t and CPU_* macros in the patch not only because it's more
generic and not depend on the implementation of the kernel (I guess the
code in Nginx now is inspired by some _early_ kernel implementation, i.e.
Linux 2.5.x), but also we will have no 32-core-limitation any more
('"worker_cpu_affinity" supports up to 32 CPU only').

I agree that the set_cpu_affinity related code in Nginx core should be
platform-independent. But the code now is _not_ (please look at those
NGX_HAVE_SCHED_SETAFFINITY macros). And I do recommend move the
implementation of ngx_set_cpu_affinity() to the os/ directory when
worker_cpu_affinity is supported on other platforms.

Thank you very much for the review :)

Regards,

-- 
Joshua Zhu
Senior Software Engineer
Server Platforms Team at Taobao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120116/d1b9afcc/attachment.html>

From nginx-forum at nginx.us  Mon Jan 16 06:14:28 2012
From: nginx-forum at nginx.us (dev)
Date: Mon, 16 Jan 2012 01:14:28 -0500
Subject: Reverse proxy for multiple websites
In-Reply-To: <20120102235905.GD67687@mdounin.ru>
References: <20120102235905.GD67687@mdounin.ru>
Message-ID: <9bfec7d5cd6030312300a3b9f019d1d1.NginxMailingListEnglish@forum.nginx.org>

hi,
Can u show ur configuration?

Dev

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220736,221282#msg-221282


From nginx-forum at nginx.us  Mon Jan 16 06:37:44 2012
From: nginx-forum at nginx.us (takigama)
Date: Mon, 16 Jan 2012 01:37:44 -0500
Subject: alias, regex and try_files am i doing it wrong, or is this a bug?
In-Reply-To: <20120113125218.GC67687@mdounin.ru>
References: <20120113125218.GC67687@mdounin.ru>
Message-ID: <c4f5eb4982cbf2df0a8db2de5c68df81.NginxMailingListEnglish@forum.nginx.org>

Maxim Dounin Wrote:
-------------------------------------------------------
> Hello!
> 
> On Thu, Jan 12, 2012 at 11:36:41PM -0500, takigama
> wrote:
> 
> > Ok, i was trying to do drupal as a subdirectory
> on my site, but i've hit
> > a problem and I think its a bug. Im using 10.04
> ubuntu, the nginx ppa
> > with 1.0.11 nginx. 
> > 
> > If i have the following config's and i try to
> retrieve the file
> > http://mysite/drupal/somefile:
> > 
> > location ~ /drupal(.*) {
> >     index index.php;
> >     alias /local_config/$1;
> >     try_files $uri @drupalrewrite;
> > }
> > 
> > The file nginx is apparently trying retrieve
> > "/local_config//somefile/drupal/somefile" rather
> then
> > "/local_config//somefile", which leads me to
> believe $uri isnt being set
> > correctly by nginx.
> 
> Yes, this looks like a bug for me.
> 
> (Internally, the problem is a bit complex, and I
> event tend to 
> think that behaviour of try_files with alias in a
> regexp location 
> is inherently undefined, but the example above
> should certainly 
> work as expected.)
> 

Thanks for confirming, was driving me a little nuts trying to work out
why it was coming up with the paths it was using.

Regards

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221138,221283#msg-221283


From agentzh at gmail.com  Mon Jan 16 07:08:05 2012
From: agentzh at gmail.com (agentzh)
Date: Mon, 16 Jan 2012 15:08:05 +0800
Subject: [ANN] ngx_headers_more module v0.16 released!
In-Reply-To: <CAB4Tn6M6t57ZTmquzLiMHb-gUnEuL9NSP+b-67fu1jMtodRp3g@mail.gmail.com>
References: <CAB4Tn6M6t57ZTmquzLiMHb-gUnEuL9NSP+b-67fu1jMtodRp3g@mail.gmail.com>
Message-ID: <CAB4Tn6MW=8Aub3JQfJLEu2RNVzvMXNWuQh8duG5bpD31jEZ4mw@mail.gmail.com>

Hello, folks!

I'm happy to announce the v0.16 release of our ngx_headers_more module
with various bug fixes accumulated in the last 6 months. You can
download the source code tarball from the download page below:

   http://github.com/agentzh/headers-more-nginx-module/tags

Special thanks go to all of our contributors and users!

Here's the complete change log for this version (as compared to the
last formal release, v0.15):

* bugfix: the on-demand handler/filter registration mechanism did not
work fully for config reload via the HUP signal.

* bugfix: when setting a multi-value response header to a single
value, the single value might be repeated on each old value.

* feature: added some debugging outputs that can be enabled by the
--with-debug option when building nginx or ngx_openresty.

* bugfix: we should set header hash using ngx_hash_key_lc, not simply to 1.

* bugfix: Setting Cache-Control response headers might not work with
other nginx output filter modules because we did not properly prepare
the r->cache_control array at the same time.

* bugfix: more_set_input_headers and more_clear_input_headers did not
handle the Accept-Encoding request headers properly. thanks ????.

* bugfix: the more_set_input_headers directive might cause invalid
memory reads because Nginx request header values must be null
terminated. thanks Maxim Dounin.

* bugfix: removing builtin headers in huge request headers with 20+
entries could result in data loss. thanks Chris Dumoulin for the patch
in github issue #6.

You may want to check out the HTML version of this change log here,
which should look better:

    http://wiki.nginx.org/HttpHeadersMoreModule#v0.16

The ngx_headers_more module allows you to add, set, or clear any
output or input header that you specify. This is an enhanced version
of the standard headers module because it provides more utilities like
resetting or clearing "builtin headers" like Content-Type,
Content-Length, and Server.

You can always get the latest source code from its project page on GitHub

   http://github.com/agentzh/headers-more-nginx-module

and the full documentation from the nginx wiki

   http://wiki.nginx.org/NginxHttpHeadersMoreModule

This module is included and enabled by default in our ngx_openresty
bundle: http://openresty.org

Enjoy!
-agentzh


From xb.zhou at qq.com  Sun Jan 15 16:31:51 2012
From: xb.zhou at qq.com (=?ISO-8859-1?B?WkhPVSBYaWFvYm8=?=)
Date: Mon, 16 Jan 2012 00:31:51 +0800
Subject: About nginx module developing
Message-ID: <tencent_276F6AA7754A46291B48C609@qq.com>

hi

I am new to nginx module developing.
I want to implement a module to deal with this demand:

1, nginx receives a request from client 
2, nginx sends two internal requests(http or protobuf etc.) to 2 backend servers
3, after the responses of the two internal requests received nginx makes the response and sends to the client

I am not sure that whether nginx can do this.
thx


------------------
Sincerely yours
                         ZHOU Xiaobo
 <br>

From ano at bestmx.ru  Mon Jan 16 10:19:41 2012
From: ano at bestmx.ru (Andrey N. Oktyabrski)
Date: Mon, 16 Jan 2012 13:19:41 +0300
Subject: About nginx module developing
In-Reply-To: <tencent_276F6AA7754A46291B48C609@qq.com>
References: <tencent_276F6AA7754A46291B48C609@qq.com>
Message-ID: <4F13F9BD.4090700@bestmx.ru>

On 15.01.12 19:31, ZHOU Xiaobo wrote:
> I am new to nginx module developing.
> I want to implement a module to deal with this demand:
>
> 1, nginx receives a request from client
> 2, nginx sends two internal requests(http or protobuf etc.) to 2 backend servers
> 3, after the responses of the two internal requests received nginx makes the response and sends to the client
>
> I am not sure that whether nginx can do this.
Maybe SSI is what you want?


From igor at sysoev.ru  Mon Jan 16 09:58:18 2012
From: igor at sysoev.ru (Igor Sysoev)
Date: Mon, 16 Jan 2012 13:58:18 +0400
Subject: Regex rewrite to root document
In-Reply-To: <026ecd3865a95af1d3751ce80a2df5fd.NginxMailingListEnglish@forum.nginx.org>
References: <026ecd3865a95af1d3751ce80a2df5fd.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120116095817.GA70724@nginx.com>

On Sun, Jan 15, 2012 at 11:35:44PM -0500, etrader wrote:
> I have a simple re-write of
> 
> rewrite ^/(.*) /post.php?id=$1 last;
> 
> This will breaks my main index page, as domain.com will also be
> redirected to "post.php?id="

location = / {
    ...
}

location ~ ^/(?<ID>.+)$ {
    fastcgi_pass   ...
    include        fastcgi_params;
    fastcgi_param  SCRIPT_FILENAME  /path/to/post/php;
    fastcgi_param  QUERY_STRING     id=$id;
}


-- 
Igor Sysoev


From mdounin at mdounin.ru  Mon Jan 16 10:51:07 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 16 Jan 2012 14:51:07 +0400
Subject: [PATCH 1 of 2] fix sched_setaffinity(2) call failure
In-Reply-To: <CAOKiq2AaxHNwB=+y1d=_x0vJE6YrZ3gC0=JGpwPPBF6OmN7bnQ@mail.gmail.com>
References: <CAOKiq2D2hwCFy3+9wFeU8ZVxtHRgU3Nrj4-UafSuYhgn8AzCOw@mail.gmail.com>
	<20120113181651.GM67687@mdounin.ru>
	<CAOKiq2AaxHNwB=+y1d=_x0vJE6YrZ3gC0=JGpwPPBF6OmN7bnQ@mail.gmail.com>
Message-ID: <20120116105107.GV67687@mdounin.ru>

Hello!

On Mon, Jan 16, 2012 at 01:52:09PM +0800, Joshua Zhu wrote:

[...]

> > diff -r e1296af53cc0 src/os/unix/ngx_process_cycle.c
> > --- a/src/os/unix/ngx_process_cycle.c   Mon Dec 26 00:00:00 2011 +0400
> > +++ b/src/os/unix/ngx_process_cycle.c   Fri Jan 13 11:58:01 2012 -0500
> > @@ -914,7 +914,10 @@
> >         ngx_log_error(NGX_LOG_NOTICE, cycle->log, 0,
> >                       "sched_setaffinity(0x%08Xl)", cpu_affinity);
> >
> > -        if (sched_setaffinity(0, 32, (cpu_set_t *) &cpu_affinity) == -1) {
> > +        if (sched_setaffinity(0, sizeof(cpu_affinity),
> > +                              (cpu_set_t *) &cpu_affinity)
> > +            == -1)
> > +        {
> >             ngx_log_error(NGX_LOG_ALERT, cycle->log, ngx_errno,
> >                           "sched_setaffinity(0x%08Xl) failed",
> > cpu_affinity);
> >         }
> >
> > Could you please test if it works for you?
> >
> 
> Yes, it works on one of my Linux box which has two cores. But I have not
> tested it on more other systems yet.

Ok, I'll commit it then, as this looks like an obvious fix.

[...]

> > While I generally agree that using cpu_set_t and CPU_* macros
> > explicitly would be better, I don't think we want to introduce
> > linux-centric code here in platform-independent code.
> >
> > Instead we probably want to leave this code generic enough to be
> > useable with other scheduler interfaces, and convert the data
> > somewhere near platform-specific whatever_setaffinity() calls if
> > needed.  (Likely unneeded for the Linux with the above patch.)
> >
> >
> We use cpu_set_t and CPU_* macros in the patch not only because it's more
> generic and not depend on the implementation of the kernel (I guess the
> code in Nginx now is inspired by some _early_ kernel implementation, i.e.
> Linux 2.5.x), but also we will have no 32-core-limitation any more
> ('"worker_cpu_affinity" supports up to 32 CPU only').

Are you facing the 32-core limit in practice?

> I agree that the set_cpu_affinity related code in Nginx core should be
> platform-independent. But the code now is _not_ (please look at those
> NGX_HAVE_SCHED_SETAFFINITY macros).

As of now, the only part which isn't portable is 
sched_setaffinity() call itself.  Other parts (i.e. config 
parsing) may be reused for other whatever_setaffinity() calls as 
soon as we add support.

> And I do recommend move the
> implementation of ngx_set_cpu_affinity() to the os/ directory when
> worker_cpu_affinity is supported on other platforms.

Yes, that's basically what I suggest also.  We want some generic 
interface, and platform-specific wrappers which do needed 
conversions/magic.

Maxim Dounin


From zhuzhaoyuan at gmail.com  Mon Jan 16 13:21:54 2012
From: zhuzhaoyuan at gmail.com (Joshua Zhu)
Date: Mon, 16 Jan 2012 21:21:54 +0800
Subject: [PATCH 1 of 2] fix sched_setaffinity(2) call failure
In-Reply-To: <20120116105107.GV67687@mdounin.ru>
References: <CAOKiq2D2hwCFy3+9wFeU8ZVxtHRgU3Nrj4-UafSuYhgn8AzCOw@mail.gmail.com>
	<20120113181651.GM67687@mdounin.ru>
	<CAOKiq2AaxHNwB=+y1d=_x0vJE6YrZ3gC0=JGpwPPBF6OmN7bnQ@mail.gmail.com>
	<20120116105107.GV67687@mdounin.ru>
Message-ID: <CAOKiq2CpdsZ82+kYe8ee1fWGLqrjtw+-9YPtdnHUu=FDAw6XKw@mail.gmail.com>

Hi,

On Mon, Jan 16, 2012 at 6:51 PM, Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
>
> On Mon, Jan 16, 2012 at 01:52:09PM +0800, Joshua Zhu wrote:
>
> [...]
>
> > > diff -r e1296af53cc0 src/os/unix/ngx_process_cycle.c
> > > --- a/src/os/unix/ngx_process_cycle.c   Mon Dec 26 00:00:00 2011 +0400
> > > +++ b/src/os/unix/ngx_process_cycle.c   Fri Jan 13 11:58:01 2012 -0500
> > > @@ -914,7 +914,10 @@
> > >         ngx_log_error(NGX_LOG_NOTICE, cycle->log, 0,
> > >                       "sched_setaffinity(0x%08Xl)", cpu_affinity);
> > >
> > > -        if (sched_setaffinity(0, 32, (cpu_set_t *) &cpu_affinity) ==
> -1) {
> > > +        if (sched_setaffinity(0, sizeof(cpu_affinity),
> > > +                              (cpu_set_t *) &cpu_affinity)
> > > +            == -1)
> > > +        {
> > >             ngx_log_error(NGX_LOG_ALERT, cycle->log, ngx_errno,
> > >                           "sched_setaffinity(0x%08Xl) failed",
> > > cpu_affinity);
> > >         }
> > >
> > > Could you please test if it works for you?
> > >
> >
> > Yes, it works on one of my Linux box which has two cores. But I have not
> > tested it on more other systems yet.
>
> Ok, I'll commit it then, as this looks like an obvious fix.
>
> [...]
>
>
OK. This fix looks good to me now. Just commit this piece of code is enough.

Regards,

-- 
Joshua Zhu
Senior Software Engineer
Server Platforms Team at Taobao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120116/95bf3f82/attachment.html>

From agentzh at gmail.com  Mon Jan 16 14:35:51 2012
From: agentzh at gmail.com (agentzh)
Date: Mon, 16 Jan 2012 22:35:51 +0800
Subject: [ANN] ngx_openresty stable version 1.0.10.24 released!
In-Reply-To: <CAB4Tn6PjohiVFgL-Cc_nA64c2iGzcxbc0QZuWt5MQ8LLGj+5Yg@mail.gmail.com>
References: <CAB4Tn6PjohiVFgL-Cc_nA64c2iGzcxbc0QZuWt5MQ8LLGj+5Yg@mail.gmail.com>
Message-ID: <CAB4Tn6PY9g2EECxYMaFDBNED54_2BUYpK=LsDv3PcX3BzfaDSQ@mail.gmail.com>

Hi, folks!

I'm happy to announce that the new stable release of ngx_openresty,
1.0.10.44, has just been kicked out of door:

  http://openresty.org/#Download

This is the second stable release of ngx_openresty that is based on the
Nginx core 1.0.10 (and the next one will be based on the Nginx core
1.0.11+).

Special thanks go to all our contributors and users to help make this
release happen :)

Here goes the complete change log for this release, as compared to the last
stable release, 1.0.10.24, released about one month ago:

   - upgraded Mike Pall's LuaJIT to 2.0.0beta9.
      - changes: http://luajit.org/changes.html
   - upgraded EchoNginxModule to 0.37.
      - bugfix: fixed two spots that we did not check null pointers
      returned by the memory allocator.
      - bugfix: attempt to fix places where ngx_time_update might not be
      compiled properly.
   - upgraded HeadersMoreNginxModule to 0.16.
      - bugfix:
more_set_input_headers<http://wiki.nginx.org/HttpHeadersMoreModule#more_set_input_headers>and
      more_clear_input_headers<http://wiki.nginx.org/HttpHeadersMoreModule#more_clear_input_headers>did
not handle the
      Accept-Encoding request headers properly. thanks ????????.
      - bugfix: removing builtin headers in huge request headers with 20+
      entries could result in data loss. thanks Chris Dumoulin for the
patch in github
      issue #6<https://github.com/agentzh/headers-more-nginx-module/issues/6>
      .
      - bugfix: the
more_set_input_headers<http://wiki.nginx.org/HttpHeadersMoreModule#more_set_headers>directive
might cause invalid memory reads because Nginx request header
      values must be null terminated. thanks Maxim Dounin.
   - upgraded LuaNginxModule to 0.4.0.
      - bugfix: ngx.flush(true) could not be used before I/O calls like
      ngx.location.capture<http://wiki.nginx.org/HttpLuaModule#ngx.location.capture>.
      this bug had appeared in v0.3.1rc34.
      - bugfix: ngx.var.VARIABLE did not evaluate to nil when the Nginx
      variable's valid flag is 0.
      - docs: various documentation improvements. thanks Nginx User.
      - bugfix: there were various places where we did not check the
      pointer returned by the memory allocator.
      - bugfix:
ngx.req.get_uri_args<http://wiki.nginx.org/HttpLuaModule#ngx.req.get_uri_args>and
      ngx.req.get_post_args<http://wiki.nginx.org/HttpLuaModule#ngx.req.get_post_args>now
only parse up to 100 arguments by default. but one can specify the
      optional argument to these two methods to specify a custom maximum number
      of args. thanks Tzury Bar Yochay for reporting this.
      - bugfix:
ngx.req.get_headers<http://wiki.nginx.org/HttpLuaModule#ngx.req.get_headers>now
only parse up to 100 request headers by default. but one can specify
      the optional argument to this method to specify a custom maximum
number of
      headers.
      - bugfix: removing builtin headers via
ngx.req.clear_header<http://wiki.nginx.org/HttpLuaModule#ngx.req.clear_header>and
its equivalent in huge request headers with 20+ entries could result
in
      data loss. thanks Chris Dumoulin for the patch in github issue
#6<https://github.com/agentzh/headers-more-nginx-module/issues/6>
      .
      - bugfix: could not compile with Nginx 1.1.12+ because Nginx 1.1.12
      changed its regex API. now we call PCRE API directly and require at least
      PCRE 8.21 for the PCRE JIT support in our ngx.re API (since PCRE 8.20
      had a bug in its JIT engine that it did not honor pcre_malloc and
      pcre_free at all).
      - bugfix:
ngx.req.set_header<http://wiki.nginx.org/HttpLuaModule#ngx.req.set_header>might
cause invalid memory reads because Nginx request header values must
      be null terminated. thanks Maxim Dounin.
      - bugfix: ngx.var.VARIABLE might evaluate to nil even if there is a
      valid value because the Nginx variable value's valid flag might not
      be initialized properly. this bad had appeared in v0.3.1rc40.
      - bugfix:
ngx.req.set_header<http://wiki.nginx.org/HttpLuaModule#ngx.req.set_header>and
      ngx.req.clear_header<http://wiki.nginx.org/HttpLuaModule#ngx.req.clear_header>did
not handle the
      Accept-Encoding request headers properly. thanks ????????.
   - upgraded EncryptedSessionNginxModule to 0.02.
      - bugfix: the -lssl option broke nginx linking when
--with-openssl=DIRis specified. thanks charlieyang for reporting this
issue.
   - upgraded SetMiscNginxModule to v0.22rc4.
      - bugfix: fixed one place that does not check the pointer returned by
      the memory allocator.
      - src: converted CRLF in the source files and test files to LF.
   - upgraded SrcacheNginxModule to v0.13rc3.
      - bugfix: fixed a regression with XssNginxModule for cache hits. this
      bug had appeared in v0.13rc1. thanks ?????? <http://weibo.com/liseen>.
      - bugfix: we did not cache the Location response header at all for 301
      /302 responses.
      - bugfix: we should not blindly cache the Accept-Ranges:
bytesresponse headers regardless of the actual current requests.
   - upgraded XssNginxModule to v0.03rc8.
      - bugfix: fixed a few debug-level log messages; the old text was
      misleading.
   - upgraded Maxim Dounin's UpstreamKeepaliveNginxModule to 0.7.
      - Bugfix: unbuffered connection might not be kept alive under Linux.
      - Bugfix: module could not be built on Windows.
      - Bugfix: module could not be built without the ngx_http_ssl_module.
      - Feature: https connections support (requires patches).
      - Bugfix: invalid connections might be cached.
      - Bugfix: the "[alert] ... open socket ... left in connection
..."messages were logged on nginx worker process gracefull exit for
each cached
      connection; the bug had appeared in version 0.3.
   - bugfix: fixed issues with relative path DIR in the
--with-openssl=DIRoption for
   ./configure.
   - bugfix: fixed compatibility of the packaging scripts on Darwin and
   *BSD. thanks nightsailer and Piotr Sikora.
   - bugfix: the install phony target did not depend on the all phony
   target in the Makefile generated by ./configure. thanks
??????<http://weibo.com/yaoweibin>for reporting this issue.
   - bugfix: the ./configure script's --add-module option did not accept
   relative path values. thanks ?????? <http://weibo.com/yaoweibin> for the
   patch.
   - bugfix: some old version of shell cp command does not support trailing
   slashes in the destination argument and could break our
./configurescript. thanks
   ?????? <http://weibo.com/yaoweibin> for reporting it.

As always, you're welcome to report bugs and feature requests either here
or directly to me :)

OpenResty (aka. ngx_openresty) is a full-fledged web application server by
bundling the standard Nginx core, lots of 3rd-party Nginx modules, as well
as most of their external dependencies.

By taking adantage of various well-designed Nginx modules,
OpenRestyeffectively turns the nginx server into a powerful web app
server, in which
the web developers can use the Lua programming language to script various
existing nginx C modules and Lua modules and construct extremely
high-performance web applications that is capable to handle 10K+
connections.

OpenResty aims to run your server-side web app completely in the Nginx
server, leveraging Nginx's event model to do non-blocking I/O not only with
the HTTP clients, but also with remote backends like MySQL, PostgreSQL,
Memcached, and Redis.

You can find more details on the homepage of ngx_openresty here:

    http://openresty.org

Have fun!
-agentzh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120116/dfa63084/attachment.html>

From agentzh at gmail.com  Mon Jan 16 14:47:32 2012
From: agentzh at gmail.com (agentzh)
Date: Mon, 16 Jan 2012 22:47:32 +0800
Subject: [ANN] ngx_openresty stable version 1.0.10.24 released!
In-Reply-To: <CAB4Tn6PY9g2EECxYMaFDBNED54_2BUYpK=LsDv3PcX3BzfaDSQ@mail.gmail.com>
References: <CAB4Tn6PjohiVFgL-Cc_nA64c2iGzcxbc0QZuWt5MQ8LLGj+5Yg@mail.gmail.com>
	<CAB4Tn6PY9g2EECxYMaFDBNED54_2BUYpK=LsDv3PcX3BzfaDSQ@mail.gmail.com>
Message-ID: <CAB4Tn6Mcjhm9dTkk7BKPJw2JwBTpne5Tc7fphEaVTBRVnGTD_Q@mail.gmail.com>

On Mon, Jan 16, 2012 at 10:35 PM, agentzh <agentzh at gmail.com> wrote:
> I'm happy to announce that the new stable release of ngx_openresty, 1.0.10.44, has just been kicked out of door:
>
> ? http://openresty.org/#Download
>

Sorry, this thread's title should be "[ANN] ngx_openresty stable
version 1.0.10.44 released!". Sorry again for the confusion :P

Best,
-agentzh


From mdounin at mdounin.ru  Mon Jan 16 16:04:40 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 16 Jan 2012 20:04:40 +0400
Subject: nginx-1.1.13
Message-ID: <20120116160440.GZ67687@mdounin.ru>

Changes with nginx 1.1.13                                        16 Jan 2012

    *) Feature: the "TLSv1.1" and "TLSv1.2" parameters of the
       "ssl_protocols" directive.

    *) Bugfix: the "limit_req" directive parameters were not inherited
       correctly; the bug had appeared in 1.1.12.

    *) Bugfix: the "proxy_redirect" directive incorrectly processed
       "Refresh" header if regular expression were used.

    *) Bugfix: the "proxy_cache_use_stale" directive with "error" parameter
       did not return answer from cache if there were no live upstreams.

    *) Bugfix: the "worker_cpu_affinity" directive might not work.

    *) Bugfix: nginx could not be built on Solaris; the bug had appeared in
       1.1.12.

    *) Bugfix: in the ngx_http_mp4_module.


Maxim Dounin


From robsonpeixoto at gmail.com  Mon Jan 16 19:56:16 2012
From: robsonpeixoto at gmail.com (Robson Roberto Souza Peixoto)
Date: Mon, 16 Jan 2012 17:56:16 -0200
Subject: Proxy with URI and params
Message-ID: <CAEA_=xJBsvzGj8tw7w0=vP097iqjGW8Mes1C8LGk=1c3LxeFzw@mail.gmail.com>

Hi,

I need to make a proxy solution. But I having some problems.

When the type www.example.com/delivery/a/b/c I need to make a proxy to
www.proxy.com/delivery/a/b/c/?var1=A&var2=B

So a did this configuration

location /delivery {
   subs_filter http://www.example.com/$uri
http://www.proxy.com/$uri/?var1=A&var2=B;
   proxy_pass http://www.proxy.com/$uri/?var1=A&var2=B;
}

Why it not working ?

Thanks and sorry for my English.

-- 
Robson Roberto Souza Peixoto
Robinho
Master in Computer Science, University of Campinas
Linux Counter #395633
IRC: robsonpeixoto
Twitter: http://twitter.com/rrspba
github: https://github.com/robsonpeixoto
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120116/a7090226/attachment.html>

From appa at perusio.net  Mon Jan 16 20:06:14 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Mon, 16 Jan 2012 20:06:14 +0000
Subject: Proxy with URI and params
In-Reply-To: <CAEA_=xJBsvzGj8tw7w0=vP097iqjGW8Mes1C8LGk=1c3LxeFzw@mail.gmail.com>
References: <CAEA_=xJBsvzGj8tw7w0=vP097iqjGW8Mes1C8LGk=1c3LxeFzw@mail.gmail.com>
Message-ID: <87fwffwgy1.wl%appa@perusio.net>

On 16 Jan 2012 19h56 WET, robsonpeixoto at gmail.com wrote:

> Hi,
>
> I need to make a proxy solution. But I having some problems.
>
> When the type www.example.com/delivery/a/b/c I need to make a proxy
> to www.proxy.com/delivery/a/b/c/?var1=A&var2=B
>
> So a did this configuration
>
> location /delivery {
> subs_filter http://www.example.com/$uri
> http://www.proxy.com/$uri/?var1=A&var2=B;
> proxy_pass http://www.proxy.com/$uri/?var1=A&var2=B;
> }
>
> Why it not working ?

Because the HTTPSubModule replaces something on the *request body* not
on the configuration.

I suppose you want something like this:

server {
   server_name www.example.com;

   location /delivery {
       proxy_pass http://www.proxy.com$uri/?var1=A&var2=B;     
   }       
}

$uri already starts with a '/'.

--- appa


From bpaquet at octo.com  Mon Jan 16 20:40:26 2012
From: bpaquet at octo.com (Bertrand Paquet)
Date: Mon, 16 Jan 2012 21:40:26 +0100
Subject: Upstream with multiple request reply
Message-ID: <CAPs0E=3_AQ7U2LtFCdu9B2b9h1tFkpQgFgwZVYRBpB1q5O=+mg@mail.gmail.com>

Hi all,

I'm trying to write an upstream which have to do multiple request / reply
with backend, for one frontend request.

Currently,
- I'm writing first request to backend in create_request handler, write
data into r->upstream->request_bufs, and returning NGX_OK
- Nginx call me on process_header. I can read data from
r->upstream->buffer. I can loop on process_header if I return NGX_AGAIN.
But I'm not able to send more data to backend. I try to add buffer in
r->upstream->request_bufs, in r->upstream->request_bufs->next, nothing work.

Anybody know how to send data to backend in the process_header callback ?

Regards,

Bertrand
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120116/a279b8db/attachment.html>

From mdounin at mdounin.ru  Mon Jan 16 21:20:20 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 17 Jan 2012 01:20:20 +0400
Subject: Upstream with multiple request reply
In-Reply-To: <CAPs0E=3_AQ7U2LtFCdu9B2b9h1tFkpQgFgwZVYRBpB1q5O=+mg@mail.gmail.com>
References: <CAPs0E=3_AQ7U2LtFCdu9B2b9h1tFkpQgFgwZVYRBpB1q5O=+mg@mail.gmail.com>
Message-ID: <20120116212020.GH67687@mdounin.ru>

Hello!

On Mon, Jan 16, 2012 at 09:40:26PM +0100, Bertrand Paquet wrote:

> Hi all,
> 
> I'm trying to write an upstream which have to do multiple request / reply
> with backend, for one frontend request.
> 
> Currently,
> - I'm writing first request to backend in create_request handler, write
> data into r->upstream->request_bufs, and returning NGX_OK
> - Nginx call me on process_header. I can read data from
> r->upstream->buffer. I can loop on process_header if I return NGX_AGAIN.
> But I'm not able to send more data to backend. I try to add buffer in
> r->upstream->request_bufs, in r->upstream->request_bufs->next, nothing work.
> 
> Anybody know how to send data to backend in the process_header callback ?

The upstream module is designed to handle "single request - single 
response" model, it's not capable of sending multiple requests to 
backend.

Maxim Dounin


From bpaquet at octo.com  Mon Jan 16 21:45:58 2012
From: bpaquet at octo.com (Bertrand Paquet)
Date: Mon, 16 Jan 2012 22:45:58 +0100
Subject: Upstream with multiple request reply
In-Reply-To: <20120116212020.GH67687@mdounin.ru>
References: <CAPs0E=3_AQ7U2LtFCdu9B2b9h1tFkpQgFgwZVYRBpB1q5O=+mg@mail.gmail.com>
	<20120116212020.GH67687@mdounin.ru>
Message-ID: <CAPs0E=1XywadkQUxc3VwMB5PVtPMZJCMcs=c5vouMHg4FLuB=Q@mail.gmail.com>

Hi,

Arg, it's not a good news :)

Do you think I can do ngx_http_subrequest to do some additional request to
back end ?
How to say to nginx to wait for subrequest before calling create_request ?

Regards,

Bertrand


On Mon, Jan 16, 2012 at 22:20, Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
>
> On Mon, Jan 16, 2012 at 09:40:26PM +0100, Bertrand Paquet wrote:
>
> > Hi all,
> >
> > I'm trying to write an upstream which have to do multiple request / reply
> > with backend, for one frontend request.
> >
> > Currently,
> > - I'm writing first request to backend in create_request handler, write
> > data into r->upstream->request_bufs, and returning NGX_OK
> > - Nginx call me on process_header. I can read data from
> > r->upstream->buffer. I can loop on process_header if I return NGX_AGAIN.
> > But I'm not able to send more data to backend. I try to add buffer in
> > r->upstream->request_bufs, in r->upstream->request_bufs->next, nothing
> work.
> >
> > Anybody know how to send data to backend in the process_header callback ?
>
> The upstream module is designed to handle "single request - single
> response" model, it's not capable of sending multiple requests to
> backend.
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120116/7a1388b4/attachment.html>

From robsonpeixoto at gmail.com  Mon Jan 16 21:54:59 2012
From: robsonpeixoto at gmail.com (Robson Roberto Souza Peixoto)
Date: Mon, 16 Jan 2012 19:54:59 -0200
Subject: Proxy with URI and params
In-Reply-To: <87fwffwgy1.wl%appa@perusio.net>
References: <CAEA_=xJBsvzGj8tw7w0=vP097iqjGW8Mes1C8LGk=1c3LxeFzw@mail.gmail.com>
	<87fwffwgy1.wl%appa@perusio.net>
Message-ID: <CAEA_=xJcx-y5HCkO3GCiD99gxVv0MnyePd84OzHX2B95Y+Q4PA@mail.gmail.com>

Thanks Antonio.

Now the problem is with the images, css and others thinks.
When the site is loaded, it try to download from www.example.com, and not
from www.proxy.com.

For example, the browser try to get the www.example.com/menu.css, but the
correct is try to download the www.proxy.com/menu.css.

How to fix it?

Thanks

On Mon, Jan 16, 2012 at 6:06 PM, Ant?nio P. P. Almeida <appa at perusio.net>wrote:

> On 16 Jan 2012 19h56 WET, robsonpeixoto at gmail.com wrote:
>
> > Hi,
> >
> > I need to make a proxy solution. But I having some problems.
> >
> > When the type www.example.com/delivery/a/b/c I need to make a proxy
> > to www.proxy.com/delivery/a/b/c/?var1=A&var2=B
> >
> > So a did this configuration
> >
> > location /delivery {
> > subs_filter http://www.example.com/$uri
> > http://www.proxy.com/$uri/?var1=A&var2=B;
> > proxy_pass http://www.proxy.com/$uri/?var1=A&var2=B;
> > }
> >
> > Why it not working ?
>
> Because the HTTPSubModule replaces something on the *request body* not
> on the configuration.
>
> I suppose you want something like this:
>
> server {
>   server_name www.example.com;
>
>   location /delivery {
>       proxy_pass http://www.proxy.com$uri/?var1=A&var2=B;
>   }
> }
>
> $uri already starts with a '/'.
>
> --- appa
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>



-- 
Robson Roberto Souza Peixoto
Robinho
Master in Computer Science, University of Campinas
Linux Counter #395633
IRC: robsonpeixoto
Twitter: http://twitter.com/rrspba
github: https://github.com/robsonpeixoto
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120116/a5425a40/attachment.html>

From mdounin at mdounin.ru  Mon Jan 16 22:17:23 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 17 Jan 2012 02:17:23 +0400
Subject: Upstream with multiple request reply
In-Reply-To: <CAPs0E=1XywadkQUxc3VwMB5PVtPMZJCMcs=c5vouMHg4FLuB=Q@mail.gmail.com>
References: <CAPs0E=3_AQ7U2LtFCdu9B2b9h1tFkpQgFgwZVYRBpB1q5O=+mg@mail.gmail.com>
	<20120116212020.GH67687@mdounin.ru>
	<CAPs0E=1XywadkQUxc3VwMB5PVtPMZJCMcs=c5vouMHg4FLuB=Q@mail.gmail.com>
Message-ID: <20120116221723.GK67687@mdounin.ru>

Hello!

On Mon, Jan 16, 2012 at 10:45:58PM +0100, Bertrand Paquet wrote:

> Hi,
> 
> Arg, it's not a good news :)
> 
> Do you think I can do ngx_http_subrequest to do some additional request to
> back end ?
> How to say to nginx to wait for subrequest before calling create_request ?

If you are ok with *independent* requests to backend (that is, you 
just need to requests two happen, but don't need to implement some 
complex protocol), than using subrequests is right way to go.

See e.g. addition filter module sources for simple subrequest 
usage example, or ssi module sources for more complex one.

Maxim Dounin

> 
> Regards,
> 
> Bertrand
> 
> 
> On Mon, Jan 16, 2012 at 22:20, Maxim Dounin <mdounin at mdounin.ru> wrote:
> 
> > Hello!
> >
> > On Mon, Jan 16, 2012 at 09:40:26PM +0100, Bertrand Paquet wrote:
> >
> > > Hi all,
> > >
> > > I'm trying to write an upstream which have to do multiple request / reply
> > > with backend, for one frontend request.
> > >
> > > Currently,
> > > - I'm writing first request to backend in create_request handler, write
> > > data into r->upstream->request_bufs, and returning NGX_OK
> > > - Nginx call me on process_header. I can read data from
> > > r->upstream->buffer. I can loop on process_header if I return NGX_AGAIN.
> > > But I'm not able to send more data to backend. I try to add buffer in
> > > r->upstream->request_bufs, in r->upstream->request_bufs->next, nothing
> > work.
> > >
> > > Anybody know how to send data to backend in the process_header callback ?
> >
> > The upstream module is designed to handle "single request - single
> > response" model, it's not capable of sending multiple requests to
> > backend.
> >
> > Maxim Dounin
> >
> > _______________________________________________
> > nginx mailing list
> > nginx at nginx.org
> > http://mailman.nginx.org/mailman/listinfo/nginx
> >

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


From appa at perusio.net  Tue Jan 17 00:41:38 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Tue, 17 Jan 2012 00:41:38 +0000
Subject: Proxy with URI and params
In-Reply-To: <CAEA_=xJcx-y5HCkO3GCiD99gxVv0MnyePd84OzHX2B95Y+Q4PA@mail.gmail.com>
References: <CAEA_=xJBsvzGj8tw7w0=vP097iqjGW8Mes1C8LGk=1c3LxeFzw@mail.gmail.com>
	<87fwffwgy1.wl%appa@perusio.net>
	<CAEA_=xJcx-y5HCkO3GCiD99gxVv0MnyePd84OzHX2B95Y+Q4PA@mail.gmail.com>
Message-ID: <87aa5nw471.wl%appa@perusio.net>

On 16 Jan 2012 21h54 WET, robsonpeixoto at gmail.com wrote:

> [1  <multipart/alternative (7bit)>]
> [1.1  <text/plain; UTF-8 (quoted-printable)>]
> Thanks Antonio.
>
> Now the problem is with the images, css and others thinks.  When the
> site is loaded, it try to download from www.example.com, and not
> from www.proxy.com.
>
> For example, the browser try to get the www.example.com/menu.css,
> but the correct is try to download the www.proxy.com/menu.css.

Then you need to be clear on what you want.

Perhaps you just and to proxy_pass to another host. Or if you want to
have other stuff like images and such try:

location /delivery {

    location ~* /delivery/[^/]*.\(?:css|js|gif|jpe?g|png)$ {
        proxy_pass http://www.proxy.com;

    }     

    proxy_pass http://www.proxy.com$uri/?var1=A&var2=B;
}

I'm just guessing. We could help you more if stated more clearly
what's your purpose.

--- appa


From appa at perusio.net  Tue Jan 17 01:04:20 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Tue, 17 Jan 2012 01:04:20 +0000
Subject: Proxy with URI and params
In-Reply-To: <87aa5nw471.wl%appa@perusio.net>
References: <CAEA_=xJBsvzGj8tw7w0=vP097iqjGW8Mes1C8LGk=1c3LxeFzw@mail.gmail.com>
	<87fwffwgy1.wl%appa@perusio.net>
	<CAEA_=xJcx-y5HCkO3GCiD99gxVv0MnyePd84OzHX2B95Y+Q4PA@mail.gmail.com>
	<87aa5nw471.wl%appa@perusio.net>
Message-ID: <878vl7w357.wl%appa@perusio.net>

On 17 Jan 2012 00h41 WET, appa at perusio.net wrote:

> On 16 Jan 2012 21h54 WET, robsonpeixoto at gmail.com wrote:
>
>> [1  <multipart/alternative (7bit)>]
>> [1.1  <text/plain; UTF-8 (quoted-printable)>]
>> Thanks Antonio.
>>
>> Now the problem is with the images, css and others thinks.  When
>> the site is loaded, it try to download from www.example.com, and
>> not from www.proxy.com.
>>
>> For example, the browser try to get the www.example.com/menu.css,
>> but the correct is try to download the www.proxy.com/menu.css.
>
> Then you need to be clear on what you want.
>
> Perhaps you just and to proxy_pass to another host. Or if you want
> to have other stuff like images and such try:
>
> location /delivery {
>
- location ~* /delivery/[^/]*.\(?:css|js|gif|jpe?g|png)$ {
+ location ~* /delivery/[^/]*\.(?:css|js|gif|jpe?g|png)$ {

--- appa


From nginx-forum at nginx.us  Tue Jan 17 03:14:55 2012
From: nginx-forum at nginx.us (bigplum)
Date: Mon, 16 Jan 2012 22:14:55 -0500
Subject: Upstream with multiple request reply
In-Reply-To: <CAPs0E=3_AQ7U2LtFCdu9B2b9h1tFkpQgFgwZVYRBpB1q5O=+mg@mail.gmail.com>
References: <CAPs0E=3_AQ7U2LtFCdu9B2b9h1tFkpQgFgwZVYRBpB1q5O=+mg@mail.gmail.com>
Message-ID: <e37a449d6e1e1274c761e67ef164eb5a.NginxMailingListEnglish@forum.nginx.org>

Hi,

This demo maybe helpful. 
https://github.com/bigplum/nginx-http-nphase-module

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221329,221339#msg-221339


From nginx-forum at nginx.us  Tue Jan 17 03:44:53 2012
From: nginx-forum at nginx.us (bigplum)
Date: Mon, 16 Jan 2012 22:44:53 -0500
Subject: test-nginx doesn't work
In-Reply-To: <980096e872cfbf4c5be7fca64b373f40.NginxMailingListEnglish@forum.nginx.org>
References: <980096e872cfbf4c5be7fca64b373f40.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <763c6346730ccdfff38fc888681fc7f4.NginxMailingListEnglish@forum.nginx.org>

It's OK now by refering to agentzh's steps. thanks all.

Additionally, maybe it's useful to set ftp mode before running "make":

export BM_UPLOAD_FTP_PASSIVE="true"

otherwise this error would occur:

*** Installing dependencies...
Can't use an undefined value as a symbol reference at
/usr/share/perl/5.10/Net/FTP/dataconn.pm line 54.
Compilation failed in require at inc/Module/Install/AutoInstall.pm line
37.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221154,221340#msg-221340


From nginx-forum at nginx.us  Tue Jan 17 05:41:13 2012
From: nginx-forum at nginx.us (cocosw)
Date: Tue, 17 Jan 2012 00:41:13 -0500
Subject: nginx-1.1.13
In-Reply-To: <20120116160440.GZ67687@mdounin.ru>
References: <20120116160440.GZ67687@mdounin.ru>
Message-ID: <a768a38568cb73e12fa1afc339b011f6.NginxMailingListEnglish@forum.nginx.org>

Good cases on 
http://www.chingtown.com/3d-phoenix-rhinestone-case-cover-protector-for-apple-iphone-4g-4s-4.html

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221313,221344#msg-221344


From nginx-forum at nginx.us  Tue Jan 17 09:13:44 2012
From: nginx-forum at nginx.us (atrus)
Date: Tue, 17 Jan 2012 04:13:44 -0500
Subject: Update/Delete cache entry.
Message-ID: <e642b2253a45b7ea15cdb031c061b0aa.NginxMailingListEnglish@forum.nginx.org>

Hi guys,

I have an entry cache with nginx :

head -n10 /tmp/nginx/f/a78359efac162eed0807e0d85761441f 
8?O????????(1O]?        ?D?
KEY: http://www.bbc.co.uk/
HTTP/1.1 200 OK
Server: Apache
Vary:
X-Country,Accept-Encoding,X-Ip_is_advertise_combined,X-Ip_is_uk_combined,
X-GeoIP
X-Cache-Action: PASS (non-cacheable)
Cache-Control: private, max-age=60
X-Cache-Age: 0
Content-Type: text/html
Content-Encoding: gzip

I'd like update this entry before expired time of cache, so I will bash
a script to delete the entry file match the URL.
I read that the cache file name is generated from md5sum of URL.
But when I use command : 

echo "http://www.bbc.co.uk/" | md5sum
5dc8a729e141d2e783c2c7e710e42fa8

The result does not match the filename cached ?! 
Is there any relation from file name and linux command md5sum ?

and is there another way to update/delete an entry of cache ?

Thank you.
Brs
Atrus@

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221355,221355#msg-221355


From francis at daoine.org  Tue Jan 17 10:02:43 2012
From: francis at daoine.org (Francis Daly)
Date: Tue, 17 Jan 2012 10:02:43 +0000
Subject: Update/Delete cache entry.
In-Reply-To: <e642b2253a45b7ea15cdb031c061b0aa.NginxMailingListEnglish@forum.nginx.org>
References: <e642b2253a45b7ea15cdb031c061b0aa.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120117100243.GB27443@craic.sysops.org>

On Tue, Jan 17, 2012 at 04:13:44AM -0500, atrus wrote:

Hi there,

Half an answer:

> I have an entry cache with nginx :
> 
> head -n10 /tmp/nginx/f/a78359efac162eed0807e0d85761441f 

> I read that the cache file name is generated from md5sum of URL.
> But when I use command : 
> 
> echo "http://www.bbc.co.uk/" | md5sum
> 5dc8a729e141d2e783c2c7e710e42fa8
> 
> The result does not match the filename cached ?! 

"echo x" generates "x newline".

Use "echo -n x", or (better) "printf %s x"

> Is there any relation from file name and linux command md5sum ?

Yes, but you must use exactly the url.

	f
-- 
Francis Daly        francis at daoine.org


From zhuzhaoyuan at gmail.com  Tue Jan 17 10:13:18 2012
From: zhuzhaoyuan at gmail.com (Joshua Zhu)
Date: Tue, 17 Jan 2012 18:13:18 +0800
Subject: [ANNOUNCE] Tengine-1.2.2
In-Reply-To: <CAOKiq2C8czdC8+XBOoxuB04cq+g-jVooz2owt=5WLPDnx2x1OQ@mail.gmail.com>
References: <CAOKiq2C8czdC8+XBOoxuB04cq+g-jVooz2owt=5WLPDnx2x1OQ@mail.gmail.com>
Message-ID: <CAOKiq2AKCHxgyTqWR9wzR+MV6P8vFkyOR_QnB-qnpKNPVWxzww@mail.gmail.com>

Hi,

On Thu, Jan 12, 2012 at 11:35 PM, Joshua Zhu <zhuzhaoyuan at gmail.com> wrote:

> Hi,
>
> We are pleased to announce that Tengine-1.2.2 has been released. And you
> can download it here: http://tengine.taobao.org/download/tengine.tar.gz
> (For those who don't know Tengine, it is a distribution of Nginx,
> including quite a few useful patches and modules developed by a group
> software engineers from Taobao)
>
> In this release, we introduced the 'input body filter' mechanism. So it's
> very handy to a developer who wants to write a Web Application Firewall. We
> merged the changes from the latest stable Nginx, i.e. Tengine is based on
> Nginx-1.0.11 now. There are also some bug fixes in this release. More
> detailed changelog is available at
> http://tengine.taobao.org/changelog.html
>

I've written a post about how to use this 'input body filter' to develop
Web Application Firewalls, taking a _naive_ module to fight hash collision
DoS attacks as an example:
http://blog.zhuzhaoyuan.com/2012/01/a-mechanism-to-help-write-web-application-firewalls-for-nginx/

Regards,

-- 
Joshua Zhu
Senior Software Engineer
Server Platforms Team at Taobao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120117/ccafa2b0/attachment.html>

From bpaquet at octo.com  Tue Jan 17 10:29:00 2012
From: bpaquet at octo.com (Bertrand Paquet)
Date: Tue, 17 Jan 2012 11:29:00 +0100
Subject: Upstream with multiple request reply
In-Reply-To: <e37a449d6e1e1274c761e67ef164eb5a.NginxMailingListEnglish@forum.nginx.org>
References: <CAPs0E=3_AQ7U2LtFCdu9B2b9h1tFkpQgFgwZVYRBpB1q5O=+mg@mail.gmail.com>
	<e37a449d6e1e1274c761e67ef164eb5a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CAPs0E=26_oHTYn1U3vWQntEa=Up5gVP4cNrEiELJFoGW9JZKKA@mail.gmail.com>

Hi all,

Finally, I do not use subrequest, I write the following code, inspired from
upstream module.
I call ngx_http_upstream_send_another_request from process_header. It's
working fine.

Feel free to comment my code, if you think I can have problem, memory
allocation, segfault or others.

Regards,

Bertrand


static void
ngx_http_upstream_send_another_request_dummy_handler(ngx_http_request_t *r,
ngx_http_upstream_t *u)
{
    ngx_log_debug0(NGX_LOG_DEBUG_HTTP, r->connection->log, 0,
                   "http upstream send another request dummy handler");
}

static ngx_int_t
ngx_http_upstream_send_another_request(ngx_http_request_t *r,
ngx_http_upstream_t *u);

static void
ngx_http_upstream_send_another_request_handler(ngx_http_request_t *r,
ngx_http_upstream_t *u)
{
    ngx_log_debug0(NGX_LOG_DEBUG_HTTP, r->connection->log, 0,
                   "http upstream send another request handler");

    ngx_http_upstream_send_another_request(r, u);
}

static ngx_int_t
ngx_http_upstream_send_another_request(ngx_http_request_t *r,
ngx_http_upstream_t *u)
{
    ngx_int_t          rc;
    ngx_connection_t  *c;

    c = u->peer.connection;

    ngx_log_debug0(NGX_LOG_DEBUG_HTTP, c->log, 0,
                   "http upstream send another request");
    //
    // if (!u->request_sent && ngx_http_upstream_test_connect(c) != NGX_OK)
{
    //     ngx_http_upstream_next(r, u, NGX_HTTP_UPSTREAM_FT_ERROR);
    //     return;
    // }

    c->log->action = "sending request to upstream";

    rc = ngx_output_chain(&u->output, u->request_sent ? NULL :
u->request_bufs);

    u->request_sent = 1;

    if (rc == NGX_ERROR) {
        return rc;
    }

    if (c->write->timer_set) {
        ngx_del_timer(c->write);
    }

    if (rc == NGX_AGAIN) {
       ngx_log_debug0(NGX_LOG_DEBUG_HTTP, c->log, 0,
                       "ngx_output_chain return NGX_AGAIN");

        u->write_event_handler =
ngx_http_upstream_send_another_request_handler;

        ngx_add_timer(c->write, u->conf->send_timeout);

        if (ngx_handle_write_event(c->write, u->conf->send_lowat) !=
NGX_OK) {
          return NGX_ERROR;
        }

        return NGX_AGAIN;
    }

    /* rc == NGX_OK */

    if (c->tcp_nopush == NGX_TCP_NOPUSH_SET) {
        if (ngx_tcp_push(c->fd) == NGX_ERROR) {
            ngx_log_error(NGX_LOG_CRIT, c->log, ngx_socket_errno,
                          ngx_tcp_push_n " failed");
            return NGX_ERROR;
        }

        c->tcp_nopush = NGX_TCP_NOPUSH_UNSET;
    }

    ngx_add_timer(c->read, u->conf->read_timeout);

// #if 1
//     if (c->read->ready) {
//
//         /* post aio operation */
//
//         /*
//          * TODO comment
//          * although we can post aio operation just in the end
//          * of ngx_http_upstream_connect() CHECK IT !!!
//          * it's better to do here because we postpone header buffer
allocation
//          */
//
//          return u->process_header(r);
//     }
// #endif

    u->write_event_handler =
ngx_http_upstream_send_another_request_dummy_handler;

    if (ngx_handle_write_event(c->write, 0) != NGX_OK) {
        return NGX_ERROR;
    }

    return NGX_AGAIN;
}

On Tue, Jan 17, 2012 at 04:14, bigplum <nginx-forum at nginx.us> wrote:

> Hi,
>
> This demo maybe helpful.
> https://github.com/bigplum/nginx-http-nphase-module
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,221329,221339#msg-221339
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120117/07d6e411/attachment-0001.html>

From robsonpeixoto at gmail.com  Tue Jan 17 11:27:17 2012
From: robsonpeixoto at gmail.com (Robson Roberto Souza Peixoto)
Date: Tue, 17 Jan 2012 09:27:17 -0200
Subject: Proxy with URI and params
In-Reply-To: <878vl7w357.wl%appa@perusio.net>
References: <CAEA_=xJBsvzGj8tw7w0=vP097iqjGW8Mes1C8LGk=1c3LxeFzw@mail.gmail.com>
	<87fwffwgy1.wl%appa@perusio.net>
	<CAEA_=xJcx-y5HCkO3GCiD99gxVv0MnyePd84OzHX2B95Y+Q4PA@mail.gmail.com>
	<87aa5nw471.wl%appa@perusio.net> <878vl7w357.wl%appa@perusio.net>
Message-ID: <CAEA_=xJsq3uciaiaqm7pn9moxTx1aPJNvLGZidw4N6tEhttkQg@mail.gmail.com>

On Mon, Jan 16, 2012 at 11:04 PM, Ant?nio P. P. Almeida <appa at perusio.net>wrote:

> On 17 Jan 2012 00h41 WET, appa at perusio.net wrote:
>
> > On 16 Jan 2012 21h54 WET, robsonpeixoto at gmail.com wrote:
> >
> >> [1  <multipart/alternative (7bit)>]
> >> [1.1  <text/plain; UTF-8 (quoted-printable)>]
> >> Thanks Antonio.
> >>
> >> Now the problem is with the images, css and others thinks.  When
> >> the site is loaded, it try to download from www.example.com, and
> >> not from www.proxy.com.
> >>
> >> For example, the browser try to get the www.example.com/menu.css,
> >> but the correct is try to download the www.proxy.com/menu.css.
> >
> > Then you need to be clear on what you want.
>
>
I'd like to navigate on http://www.proxy.com/delivery/.../.../.../ using my
domain http://www.example.com/.

So the user will navigate on www.example.com and always show on URL the '
www.example.com/delivery/../../..'

It's is possible?


> > Perhaps you just and to proxy_pass to another host. Or if you want
> > to have other stuff like images and such try:
> >
> > location /delivery {
> >
> - location ~* /delivery/[^/]*.\(?:css|js|gif|jpe?g|png)$ {
> + location ~* /delivery/[^/]*\.(?:css|js|gif|jpe?g|png)$ {
>
> --- appa
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>



-- 
Robson Roberto Souza Peixoto
Robinho
Master in Computer Science, University of Campinas
Linux Counter #395633
IRC: robsonpeixoto
Twitter: http://twitter.com/rrspba
github: https://github.com/robsonpeixoto
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120117/b612b0c2/attachment.html>

From appa at perusio.net  Tue Jan 17 12:50:49 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Tue, 17 Jan 2012 12:50:49 +0000
Subject: Proxy with URI and params
In-Reply-To: <CAEA_=xJsq3uciaiaqm7pn9moxTx1aPJNvLGZidw4N6tEhttkQg@mail.gmail.com>
References: <CAEA_=xJBsvzGj8tw7w0=vP097iqjGW8Mes1C8LGk=1c3LxeFzw@mail.gmail.com>
	<87fwffwgy1.wl%appa@perusio.net>
	<CAEA_=xJcx-y5HCkO3GCiD99gxVv0MnyePd84OzHX2B95Y+Q4PA@mail.gmail.com>
	<87aa5nw471.wl%appa@perusio.net> <878vl7w357.wl%appa@perusio.net>
	<CAEA_=xJsq3uciaiaqm7pn9moxTx1aPJNvLGZidw4N6tEhttkQg@mail.gmail.com>
Message-ID: <877h0qwl06.wl%appa@perusio.net>

On 17 Jan 2012 11h27 WET, robsonpeixoto at gmail.com wrote:

> I'd like to navigate on http://www.proxy.com/delivery/.../.../.../
> using my domain http://www.example.com/.
>
> So the user will navigate on www.example.com and always show on URL
> the ' www.example.com/delivery/../../..'
>
> It's is possible?

Yes. Add the proxy_redirect directive:

proxy_redirect http://www.proxy.com/delivery/ http://www.example.com/;

See: http://wiki.nginx.org/HttpProxyModule#proxy_redirect

--- appa


From appa at perusio.net  Tue Jan 17 12:56:45 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Tue, 17 Jan 2012 12:56:45 +0000
Subject: Update/Delete cache entry.
In-Reply-To: <e642b2253a45b7ea15cdb031c061b0aa.NginxMailingListEnglish@forum.nginx.org>
References: <e642b2253a45b7ea15cdb031c061b0aa.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <8762gawkqa.wl%appa@perusio.net>

On 17 Jan 2012 09h13 WET, nginx-forum at nginx.us wrote:

> Hi guys,
>
> I have an entry cache with nginx :

AFAIK there are two ways to purge the Nginx cache. 

1. Using grep throughout the cache directories to find and then
   remove the file.

2. Using an HTTP "method" PURGE.

For the first this might be helpful: 
https://github.com/perusio/nginx-cache-purge

and for the later use this: 
https://github.com/FRiCKLE/ngx_cache_purge

--- appa


From nginx-forum at nginx.us  Tue Jan 17 14:38:04 2012
From: nginx-forum at nginx.us (atrus)
Date: Tue, 17 Jan 2012 09:38:04 -0500
Subject: Update/Delete cache entry.
In-Reply-To: <e642b2253a45b7ea15cdb031c061b0aa.NginxMailingListEnglish@forum.nginx.org>
References: <e642b2253a45b7ea15cdb031c061b0aa.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <1ca6317d14bd4986fa348da67560bccc.NginxMailingListEnglish@forum.nginx.org>

Thanks guys. 
I'll have a try and result later. 

Brs. 
Atrus@

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221355,221365#msg-221365


From lists at ruby-forum.com  Tue Jan 17 17:15:50 2012
From: lists at ruby-forum.com (Jeff J.)
Date: Tue, 17 Jan 2012 18:15:50 +0100
Subject: Limit robots
In-Reply-To: <6300b8a589bb337813f7df86d393d332.NginxMailingListEnglish@forum.nginx.org>
References: <6300b8a589bb337813f7df86d393d332.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <2a8f95129b4ee5c94ed132fe1cae1f90@ruby-forum.com>

Your solution works great for rate limiting,  but I need the requests to 
still be 200 return code when they are not limited.  Any ideas?

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


From kworthington at gmail.com  Tue Jan 17 18:47:46 2012
From: kworthington at gmail.com (Kevin Worthington)
Date: Tue, 17 Jan 2012 13:47:46 -0500
Subject: nginx-1.1.13
In-Reply-To: <20120116160440.GZ67687@mdounin.ru>
References: <20120116160440.GZ67687@mdounin.ru>
Message-ID: <CAGo79UWu6EB6vidxMwkLLNAhsjaf7DFx+Xt74tft_Ybk8kbuaA@mail.gmail.com>

Hello Nginx Users,

Just released: Nginx 1.1.13 For Windows http://goo.gl/dHNTJ (32-bit
and 64-bit versions)

These versions are to support legacy users who are already using
Cygwin based builds of Nginx. Official Windows binaries are at
nginx.org

Thank you,
Kevin
--
Kevin Worthington
kworthington *@~ #gmail}  [dot) {com]
http://www.kevinworthington.com/


On Mon, Jan 16, 2012 at 11:04 AM, Maxim Dounin <mdounin at mdounin.ru> wrote:
> Changes with nginx 1.1.13 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?16 Jan 2012
>
> ? ?*) Feature: the "TLSv1.1" and "TLSv1.2" parameters of the
> ? ? ? "ssl_protocols" directive.
>
> ? ?*) Bugfix: the "limit_req" directive parameters were not inherited
> ? ? ? correctly; the bug had appeared in 1.1.12.
>
> ? ?*) Bugfix: the "proxy_redirect" directive incorrectly processed
> ? ? ? "Refresh" header if regular expression were used.
>
> ? ?*) Bugfix: the "proxy_cache_use_stale" directive with "error" parameter
> ? ? ? did not return answer from cache if there were no live upstreams.
>
> ? ?*) Bugfix: the "worker_cpu_affinity" directive might not work.
>
> ? ?*) Bugfix: nginx could not be built on Solaris; the bug had appeared in
> ? ? ? 1.1.12.
>
> ? ?*) Bugfix: in the ngx_http_mp4_module.
>
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From appa at perusio.net  Tue Jan 17 19:27:44 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Tue, 17 Jan 2012 19:27:44 +0000
Subject: Limit robots
In-Reply-To: <2a8f95129b4ee5c94ed132fe1cae1f90@ruby-forum.com>
References: <6300b8a589bb337813f7df86d393d332.NginxMailingListEnglish@forum.nginx.org>
	<2a8f95129b4ee5c94ed132fe1cae1f90@ruby-forum.com>
Message-ID: <87r4yyuo27.wl%appa@perusio.net>

On 17 Jan 2012 17h15 WET, lists at ruby-forum.com wrote:

> Your solution works great for rate limiting, but I need the requests
> to still be 200 return code when they are not limited.  Any ideas?

Yes. Try:

- error_page 418 @bots;
+ error_page 418 =200 @bots;

I.e.:

location / {
    error_page 418 =200 @bots;

    if ($is_bot) {
      return 418;  
    }
    ...
}

location @bots {
   limit_req zone=useragenttrack burst=100 nodelay;  
   ...
}

--- appa


From lists at ruby-forum.com  Tue Jan 17 20:07:14 2012
From: lists at ruby-forum.com (Jeff J.)
Date: Tue, 17 Jan 2012 21:07:14 +0100
Subject: Limit robots
In-Reply-To: <6300b8a589bb337813f7df86d393d332.NginxMailingListEnglish@forum.nginx.org>
References: <6300b8a589bb337813f7df86d393d332.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <ca4c48934369eefccc1a43f81324e992@ruby-forum.com>

Thanks, that worked great for me.

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


From henry.ward at secondsightco.com  Wed Jan 18 00:35:40 2012
From: henry.ward at secondsightco.com (Henry Ward)
Date: Tue, 17 Jan 2012 16:35:40 -0800
Subject: Ningx/Django Bad Request Error
Message-ID: <86A60AE8-FAFB-451F-BDB9-24EEF89B9264@secondsightco.com>

Hello Nginx Group.

I am new to Ngnix and installing my first Nginx/Django server.  I've been struggling with a Bad Request 400 Error for two days now and I have nowhere else to go.  Here is the problem:

I have Nginx configured with django.  My default config file is below.  I am have a basic django-piston REST service setup that works correctly with django's runserver.  The following two commands work perfectly with django's runserver http server:

url = 'http://localhost/receipts'
# GET receipts data
req = urllib2.Request(url)		
f = urllib2.urlopen(req)

# POST Receipts data with json data
data = <json data...>
req = urllib2.Request(url, data=data, headers={'Content-Type': 'application/json'}) 	
f = urllib2.urlopen(req)

When I run the same requests against my nginx installation the GET function works just fine.   However, the POST request returns a HTTP Error 400: BAD REQUEST error.  

It is exactly the same code and I know this is working with the django runserver.  It must be some configuration in Nginx that is regecting the POST request.  I get that error for any Content-Type.  

I have looked everywhere and I cannot find anything about allowing this post request.   I hope someone here can help nudge me in the right direction.  My Nginx config file settings is below.  I have modified many variables like buffer sizes and body sizes and it is not that.  The data I'm posting is less than 8k.  

Any ideas?

----------------------------

server {
        include mime.types;
        default_type application/json;

        client_body_buffer_size 1200K;
        client_header_buffer_size 1200K;
        client_max_body_size 2m;
        large_client_header_buffers 2 1200K;

        listen   80; ## listen for ipv4; this line is default and implied
        #listen   [::]:80 default ipv6only=on; ## listen for ipv6


        # Make site accessible from http://localhost/
        server_name localhost;

        # Log information
        access_log /home/henry/logs/access.log;
        error_log /home/henry/logs/error.log;

        root /home/henry/webservice/public;

        location /site_media
        {
                root /path/to/yoursite.com/public;
        }

        location / {
                        # host and port to fastcgi server
                        fastcgi_pass 127.0.0.1:8081;
                        fastcgi_param PATH_INFO $fastcgi_script_name;
                        fastcgi_param REQUEST_METHOD $request_method;
                        fastcgi_param QUERY_STRING $query_string;
                        fastcgi_param SERVER_NAME $server_name;
                        fastcgi_param SERVER_PORT $server_port;
                        fastcgi_param SERVER_PROTOCOL $server_protocol;
                        fastcgi_param CONTENT_TYPE $content_type;
                        fastcgi_param CONTENT_LENGTH $content_length;
                        fastcgi_pass_header Authorization;
                        fastcgi_intercept_errors off;
        }

        location /doc {
                root /usr/share;
                autoindex on;
                allow 127.0.0.1;
                deny all;
        }

        location /images {
                root /usr/share;
                autoindex off;
        }

}



From nginxlist at serverphorums.com  Tue Jan 17 18:01:01 2012
From: nginxlist at serverphorums.com (nginxlist at serverphorums.com)
Date: Tue, 17 Jan 2012 19:01:01 +0100 (CET)
Subject: nginx keep using more and more memory
Message-ID: <8d53edb63ae48f2c1bb1b904bde571e0.Nginx@www.serverphorums.com>

Hi,

We are using nginx to proxy to 4 back end servers. And we terminate SSL at nginx. We use nginx in a somewhat special way in that we make 1 HTTP GET and the back-end servers give a very very long (essentially never-ending) response.

trouble is that i notice that the nginx worker processes keep using up more and more memory. When the connections to nginx terminate, everything seems to be cleaned up and the memory usage of the worker processes drops to normal.

my ssl config looks like this:

# HTTPS server
    server {
        server_name  myserver;
        listen       myserver:4443;
        ssl on;

        ssl_certificate      /myserver.pem;
        ssl_certificate_key  /myserver.pem;

        proxy_ssl_session_reuse off;

        ssl_protocols  SSLv3 TLSv1;
        ssl_ciphers  HIGH:!ADH:!MD5;
        ssl_prefer_server_ciphers   on;

        proxy_buffering off;

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


        location /SERVER_1/ {
            rewrite /SERVER_1/(.*) /$1 break;
            proxy_set_header  X-Real-IP  $remote_addr;
            proxy_pass http://localhost:8080;
        }
}

i am running nginx/1.0.11

I can reproduce the problem each time with a test setup of:

1 nginx and a special servlet that only dishes out a constant stream of serialized java objects. If you are interested, i can provide also this servlet. then using about 50 long-running HTTP GETs (simulating 100 clients), i can always get more memory usage in nginx.

does anyone know if i misconfigured my nginx?

---
posted at http://www.serverphorums.com
http://www.serverphorums.com/read.php?5,431797,431797#msg-431797


From delta.yeh at gmail.com  Wed Jan 18 09:27:01 2012
From: delta.yeh at gmail.com (Delta Yeh)
Date: Wed, 18 Jan 2012 17:27:01 +0800
Subject: nginx keep using more and more memory
In-Reply-To: <8d53edb63ae48f2c1bb1b904bde571e0.Nginx@www.serverphorums.com>
References: <8d53edb63ae48f2c1bb1b904bde571e0.Nginx@www.serverphorums.com>
Message-ID: <CADKMG5345YjGF97S+A6otQ5U1yOpm0wZekgiz1uoBYJgwhJxWg@mail.gmail.com>

nginx use memory pool for each http request.
If the request is never-ending, more and more memory will be allocated.

2012/1/18  <nginxlist at serverphorums.com>:
> Hi,
>
> We are using nginx to proxy to 4 back end servers. And we terminate SSL at nginx. We use nginx in a somewhat special way in that we make 1 HTTP GET and the back-end servers give a very very long (essentially never-ending) response.
>
> trouble is that i notice that the nginx worker processes keep using up more and more memory. When the connections to nginx terminate, everything seems to be cleaned up and the memory usage of the worker processes drops to normal.
>
> my ssl config looks like this:
>
> # HTTPS server
> ? ?server {
> ? ? ? ?server_name ?myserver;
> ? ? ? ?listen ? ? ? myserver:4443;
> ? ? ? ?ssl on;
>
> ? ? ? ?ssl_certificate ? ? ?/myserver.pem;
> ? ? ? ?ssl_certificate_key ?/myserver.pem;
>
> ? ? ? ?proxy_ssl_session_reuse off;
>
> ? ? ? ?ssl_protocols ?SSLv3 TLSv1;
> ? ? ? ?ssl_ciphers ?HIGH:!ADH:!MD5;
> ? ? ? ?ssl_prefer_server_ciphers ? on;
>
> ? ? ? ?proxy_buffering off;
>
> ? ? ? ?location / {
> ? ? ? ? ? ?root ? /www/server/html;
> ? ? ? ? ? ?index ?index.html index.htm;
> ? ? ? ?}
>
>
> ? ? ? ?location /SERVER_1/ {
> ? ? ? ? ? ?rewrite /SERVER_1/(.*) /$1 break;
> ? ? ? ? ? ?proxy_set_header ?X-Real-IP ?$remote_addr;
> ? ? ? ? ? ?proxy_pass http://localhost:8080;
> ? ? ? ?}
> }
>
> i am running nginx/1.0.11
>
> I can reproduce the problem each time with a test setup of:
>
> 1 nginx and a special servlet that only dishes out a constant stream of serialized java objects. If you are interested, i can provide also this servlet. then using about 50 long-running HTTP GETs (simulating 100 clients), i can always get more memory usage in nginx.
>
> does anyone know if i misconfigured my nginx?
>
> ---
> posted at http://www.serverphorums.com
> http://www.serverphorums.com/read.php?5,431797,431797#msg-431797
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From grangier at gmail.com  Wed Jan 18 09:37:43 2012
From: grangier at gmail.com (xgdlm)
Date: Wed, 18 Jan 2012 10:37:43 +0100
Subject: Ningx/Django Bad Request Error
In-Reply-To: <86A60AE8-FAFB-451F-BDB9-24EEF89B9264@secondsightco.com>
References: <86A60AE8-FAFB-451F-BDB9-24EEF89B9264@secondsightco.com>
Message-ID: <AFCD5FAA-FE74-4EFE-B6EA-8722627496B4@gmail.com>

Hello,


Le 18 janv. 2012 ? 01:35, Henry Ward a ?crit :

> Any ideas?

With this much information no.

Imho running django using fcgi is something you don't want to do.
I would recommend using http://gunicorn.org as a wsgi server (or uwgi module)

Take a look here too:
http://stackoverflow.com/questions/5706230/nginx-gunicorn-post-request-error

In the case you still want to use the old fcgi way, you should provide use with debug log on both site (nginx and application server).

Regrds,

xav

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120118/92384592/attachment.html>

From ne at vbart.ru  Wed Jan 18 09:50:51 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Wed, 18 Jan 2012 13:50:51 +0400
Subject: Ningx/Django Bad Request Error
In-Reply-To: <86A60AE8-FAFB-451F-BDB9-24EEF89B9264@secondsightco.com>
References: <86A60AE8-FAFB-451F-BDB9-24EEF89B9264@secondsightco.com>
Message-ID: <201201181350.51503.ne@vbart.ru>

On Wednesday 18 January 2012 04:35:40 Henry Ward wrote:
[...]
> 
> I have looked everywhere and I cannot find anything about allowing this
> post request.   I hope someone here can help nudge me in the right
> direction.  My Nginx config file settings is below.  I have modified many
> variables like buffer sizes and body sizes and it is not that.  The data
> I'm posting is less than 8k.
> 
> Any ideas?

Could you provide debug log with your POST request?
http://nginx.org/en/docs/debugging_log.html

> ----------------------------
> 
> server {
>         include mime.types;
>         default_type application/json;
> 
>         client_body_buffer_size 1200K;
>         client_header_buffer_size 1200K;

It's a lot!


 wbr, Valentin V. Bartenev


From ne at vbart.ru  Wed Jan 18 10:00:16 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Wed, 18 Jan 2012 14:00:16 +0400
Subject: nginx keep using more and more memory
In-Reply-To: <8d53edb63ae48f2c1bb1b904bde571e0.Nginx@www.serverphorums.com>
References: <8d53edb63ae48f2c1bb1b904bde571e0.Nginx@www.serverphorums.com>
Message-ID: <201201181400.16393.ne@vbart.ru>

On Tuesday 17 January 2012 22:01:01 nginxlist at serverphorums.com wrote:
> Hi,
> 
> We are using nginx to proxy to 4 back end servers. And we terminate SSL at
> nginx. We use nginx in a somewhat special way in that we make 1 HTTP GET
> and the back-end servers give a very very long (essentially never-ending)
> response.
> 
> trouble is that i notice that the nginx worker processes keep using up more
> and more memory. When the connections to nginx terminate, everything seems
> to be cleaned up and the memory usage of the worker processes drops to
> normal.
> 
[...]

What openssl version do you use?

 wbr, Valentin V. Bartenev


From ne at vbart.ru  Wed Jan 18 10:18:25 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Wed, 18 Jan 2012 14:18:25 +0400
Subject: nginx keep using more and more memory
In-Reply-To: <CADKMG5345YjGF97S+A6otQ5U1yOpm0wZekgiz1uoBYJgwhJxWg@mail.gmail.com>
References: <8d53edb63ae48f2c1bb1b904bde571e0.Nginx@www.serverphorums.com>
	<CADKMG5345YjGF97S+A6otQ5U1yOpm0wZekgiz1uoBYJgwhJxWg@mail.gmail.com>
Message-ID: <201201181418.25888.ne@vbart.ru>

On Wednesday 18 January 2012 13:27:01 Delta Yeh wrote:
> nginx use memory pool for each http request.
> If the request is never-ending, more and more memory will be allocated.
> 

No. It's not true.

Please, see the following documentation:
http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_buffering
http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_buffers

also:
http://wiki.nginx.org/HttpProxyModule#proxy_max_temp_file_size

 wbr, Valentin V. Bartenev


From mdounin at mdounin.ru  Wed Jan 18 10:31:55 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Wed, 18 Jan 2012 14:31:55 +0400
Subject: nginx keep using more and more memory
In-Reply-To: <8d53edb63ae48f2c1bb1b904bde571e0.Nginx@www.serverphorums.com>
References: <8d53edb63ae48f2c1bb1b904bde571e0.Nginx@www.serverphorums.com>
Message-ID: <20120118103155.GR67687@mdounin.ru>

Hello!

On Tue, Jan 17, 2012 at 07:01:01PM +0100, nginxlist at serverphorums.com wrote:

> We are using nginx to proxy to 4 back end servers. And we 
> terminate SSL at nginx. We use nginx in a somewhat special way 
> in that we make 1 HTTP GET and the back-end servers give a very 
> very long (essentially never-ending) response.
> 
> trouble is that i notice that the nginx worker processes keep 
> using up more and more memory. When the connections to nginx 
> terminate, everything seems to be cleaned up and the memory 
> usage of the worker processes drops to normal.

[...]

> i am running nginx/1.0.11

Please try 1.1.x.  There is one known problem in 1.0.x which 
causes it to allocate about 2 pointers per buffer sent.  This is 
known to be noticeable in long-running response case.

The problem was fixed in 1.1.4, but unfortunately it's impossible 
to merge the fix into 1.0.x due to required API change.

Maxim Dounin


From saqib.rasul at gmail.com  Wed Jan 18 10:56:45 2012
From: saqib.rasul at gmail.com (Saqib Rasul)
Date: Wed, 18 Jan 2012 11:56:45 +0100
Subject: nginx keep using more and more memory
In-Reply-To: <20120118103155.GR67687@mdounin.ru>
References: <8d53edb63ae48f2c1bb1b904bde571e0.Nginx@www.serverphorums.com>
	<20120118103155.GR67687@mdounin.ru>
Message-ID: <CANRKtT0xV89uoOm_HDvRqQHvbUfkxQaDKrZi+o9f1YRnD6K5kw@mail.gmail.com>

Hi,

thanks for the answer.

regarding your tip about proxy_buffers, i dont think its that because
i have disabled proxy_buffering.

I will try out my tests with 1.1 version of nginx and let you know.

regards,

Saqib

On Wed, Jan 18, 2012 at 11:31 AM, Maxim Dounin <mdounin at mdounin.ru> wrote:
> Hello!
>
> On Tue, Jan 17, 2012 at 07:01:01PM +0100, nginxlist at serverphorums.com wrote:
>
>> We are using nginx to proxy to 4 back end servers. And we
>> terminate SSL at nginx. We use nginx in a somewhat special way
>> in that we make 1 HTTP GET and the back-end servers give a very
>> very long (essentially never-ending) response.
>>
>> trouble is that i notice that the nginx worker processes keep
>> using up more and more memory. When the connections to nginx
>> terminate, everything seems to be cleaned up and the memory
>> usage of the worker processes drops to normal.
>
> [...]
>
>> i am running nginx/1.0.11
>
> Please try 1.1.x. ?There is one known problem in 1.0.x which
> causes it to allocate about 2 pointers per buffer sent. ?This is
> known to be noticeable in long-running response case.
>
> The problem was fixed in 1.1.4, but unfortunately it's impossible
> to merge the fix into 1.0.x due to required API change.
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From saqib.rasul at gmail.com  Wed Jan 18 12:33:22 2012
From: saqib.rasul at gmail.com (Saqib Rasul)
Date: Wed, 18 Jan 2012 13:33:22 +0100
Subject: nginx keep using more and more memory
In-Reply-To: <CANRKtT0xV89uoOm_HDvRqQHvbUfkxQaDKrZi+o9f1YRnD6K5kw@mail.gmail.com>
References: <8d53edb63ae48f2c1bb1b904bde571e0.Nginx@www.serverphorums.com>
	<20120118103155.GR67687@mdounin.ru>
	<CANRKtT0xV89uoOm_HDvRqQHvbUfkxQaDKrZi+o9f1YRnD6K5kw@mail.gmail.com>
Message-ID: <CANRKtT21n+XnX1WDDFNaZfcqm2+j-kSFH3rz9gnBLdApGEihPw@mail.gmail.com>

Also do you know or can give an estimate of when the 1.1 version will go stable?

regards,

Saqib

On Wed, Jan 18, 2012 at 11:56 AM, Saqib Rasul <saqib.rasul at gmail.com> wrote:
> Hi,
>
> thanks for the answer.
>
> regarding your tip about proxy_buffers, i dont think its that because
> i have disabled proxy_buffering.
>
> I will try out my tests with 1.1 version of nginx and let you know.
>
> regards,
>
> Saqib
>
> On Wed, Jan 18, 2012 at 11:31 AM, Maxim Dounin <mdounin at mdounin.ru> wrote:
>> Hello!
>>
>> On Tue, Jan 17, 2012 at 07:01:01PM +0100, nginxlist at serverphorums.com wrote:
>>
>>> We are using nginx to proxy to 4 back end servers. And we
>>> terminate SSL at nginx. We use nginx in a somewhat special way
>>> in that we make 1 HTTP GET and the back-end servers give a very
>>> very long (essentially never-ending) response.
>>>
>>> trouble is that i notice that the nginx worker processes keep
>>> using up more and more memory. When the connections to nginx
>>> terminate, everything seems to be cleaned up and the memory
>>> usage of the worker processes drops to normal.
>>
>> [...]
>>
>>> i am running nginx/1.0.11
>>
>> Please try 1.1.x. ?There is one known problem in 1.0.x which
>> causes it to allocate about 2 pointers per buffer sent. ?This is
>> known to be noticeable in long-running response case.
>>
>> The problem was fixed in 1.1.4, but unfortunately it's impossible
>> to merge the fix into 1.0.x due to required API change.
>>
>> Maxim Dounin
>>
>> _______________________________________________
>> nginx mailing list
>> nginx at nginx.org
>> http://mailman.nginx.org/mailman/listinfo/nginx


From ne at vbart.ru  Wed Jan 18 12:52:23 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Wed, 18 Jan 2012 16:52:23 +0400
Subject: nginx keep using more and more memory
In-Reply-To: <CANRKtT21n+XnX1WDDFNaZfcqm2+j-kSFH3rz9gnBLdApGEihPw@mail.gmail.com>
References: <8d53edb63ae48f2c1bb1b904bde571e0.Nginx@www.serverphorums.com>
	<CANRKtT0xV89uoOm_HDvRqQHvbUfkxQaDKrZi+o9f1YRnD6K5kw@mail.gmail.com>
	<CANRKtT21n+XnX1WDDFNaZfcqm2+j-kSFH3rz9gnBLdApGEihPw@mail.gmail.com>
Message-ID: <201201181652.23775.ne@vbart.ru>

On Wednesday 18 January 2012 16:33:22 Saqib Rasul wrote:
> Also do you know or can give an estimate of when the 1.1 version will go
> stable?
> 

http://trac.nginx.org/nginx/milestone/1.1.17

FYI, nginx "devel" vs. "stable" difference mainly is about API and behavior 
stability. Both branches are reliable enough to use in production.

 wbr, Valentin V. Bartenev


From nginx-forum at nginx.us  Wed Jan 18 16:35:24 2012
From: nginx-forum at nginx.us (iberkner)
Date: Wed, 18 Jan 2012 11:35:24 -0500
Subject: Redirect ALL requests to single PHP file
Message-ID: <73220e73584a4118644bf6d0da8c7582.NginxMailingListEnglish@forum.nginx.org>

Hi All,

We'd like to redirect all requests to a single PHP file.  The
configuration below is what we have but I'm concerned that it would
cause an "infinite" loop and possibly other problems.  We'd like to
redirect ANY request to api.php regardless if the file exists or not. 
Even a request to an existing static file should rewrite to the
"api.php" file.

Is this the correct way? is there a better way?

Thanks

        server
        {
                listen #.#.#.#:80;
                server_name api.domain.com;

                rewrite ^(.*)$ /api.php?uri=$1 last;

                location ~ \.php$
                {
                        include fcgi;
                        fastcgi_pass 127.0.0.1:9000;
                }
        }

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221398,221398#msg-221398


From appa at perusio.net  Wed Jan 18 16:45:14 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Wed, 18 Jan 2012 16:45:14 +0000
Subject: Redirect ALL requests to single PHP file
In-Reply-To: <73220e73584a4118644bf6d0da8c7582.NginxMailingListEnglish@forum.nginx.org>
References: <73220e73584a4118644bf6d0da8c7582.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87aa5lklid.wl%appa@perusio.net>

On 18 Jan 2012 16h35 WET, nginx-forum at nginx.us wrote:

> Hi All,
>
> We'd like to redirect all requests to a single PHP file.  The
> configuration below is what we have but I'm concerned that it would
> cause an "infinite" loop and possibly other problems.  We'd like to
> redirect ANY request to api.php regardless if the file exists or
> not.  Even a request to an existing static file should rewrite to
> the "api.php" file.
>
> Is this the correct way? is there a better way?

Try:

location = /api.php {
    internal; # comment this out to allow direct access to api.php.
    include fcgi;
    fastcgi_pass 127.0.0.1:9000;      
}

location / {
    return 302 /api.php?uri=$request_uri;     
}

--- appa


From mdounin at mdounin.ru  Wed Jan 18 16:48:25 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Wed, 18 Jan 2012 20:48:25 +0400
Subject: Redirect ALL requests to single PHP file
In-Reply-To: <73220e73584a4118644bf6d0da8c7582.NginxMailingListEnglish@forum.nginx.org>
References: <73220e73584a4118644bf6d0da8c7582.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120118164825.GA67687@mdounin.ru>

Hello!

On Wed, Jan 18, 2012 at 11:35:24AM -0500, iberkner wrote:

> Hi All,
> 
> We'd like to redirect all requests to a single PHP file.  The
> configuration below is what we have but I'm concerned that it would
> cause an "infinite" loop and possibly other problems.  We'd like to
> redirect ANY request to api.php regardless if the file exists or not. 
> Even a request to an existing static file should rewrite to the
> "api.php" file.
> 
> Is this the correct way? is there a better way?
> 
> Thanks
> 
>         server
>         {
>                 listen #.#.#.#:80;
>                 server_name api.domain.com;
> 
>                 rewrite ^(.*)$ /api.php?uri=$1 last;
> 
>                 location ~ \.php$
>                 {
>                         include fcgi;
>                         fastcgi_pass 127.0.0.1:9000;
>                 }
>         }

server {
    listen ...
    server_name api.domain.com;

    location / {
        fastcgi_pass 127.0.0.1:9000;
        fastcgi_param REQUEST_FILENAME $document_root/api.php;
        include fastcgi_params;
    }
}

Maxim Dounin


From henry.ward at secondsightco.com  Wed Jan 18 16:58:48 2012
From: henry.ward at secondsightco.com (Henry Ward)
Date: Wed, 18 Jan 2012 08:58:48 -0800
Subject: Ningx/Django Bad Request Error
In-Reply-To: <86A60AE8-FAFB-451F-BDB9-24EEF89B9264@secondsightco.com>
References: <86A60AE8-FAFB-451F-BDB9-24EEF89B9264@secondsightco.com>
Message-ID: <7C5A503B-B113-47C6-AB83-26726A9FDB50@secondsightco.com>

Thanks Xav and Valentin for the quick response.  Sorry - I have to copy your responses because I was subscribed on digest rather than real-time email.  But I am now on the normal mailing list.

Error Log
The error log shows nothing.  Literally.  The access_log shows the access url and the 400 error.  But the error log does not.  I know that is hard to believe its similar to what is said here:

http://www.ruby-forum.com/topic/2573586

In that the also had the same problem that Bad Request nothing shows in the error log.  I thought it might be a similar issue but I don't have any spaces in python urllib2 header so I don't think that is the issue.

I have also created application logging in Django and have "GOT HERE" log statement in the first line of the view that handles the POST request..  That line never gets called so it appears Nginx is kicking out the request before it ever passes to the django app.

This is what makes so irritatingly difficult to debug.  It is also I feel it some type of permission error or setting that does not allow POST requests.  I've also read here (https://calomel.org/nginx.html) that Nginx by default only allows GET requests but I couldn't find anywhere where the is verified nor how to enable that.

I'm stumped.

WSGI
Xav - I am happy try gunicorn.org as a wsgi server instead of fcgi.  Particularly if there isn't a reasonable solution to this problem.  Again - I'm now to this.  Can you give a quick background on why you don't recommend fcgi and do recommend gunicorn?  (current. problem excluded ;-)

Henry


On Jan 17, 2012, at 4:35 PM, Henry Ward wrote:

> Hello Nginx Group.
> 
> I am new to Ngnix and installing my first Nginx/Django server.  I've been struggling with a Bad Request 400 Error for two days now and I have nowhere else to go.  Here is the problem:
> 
> I have Nginx configured with django.  My default config file is below.  I am have a basic django-piston REST service setup that works correctly with django's runserver.  The following two commands work perfectly with django's runserver http server:
> 
> url = 'http://localhost/receipts'
> # GET receipts data
> req = urllib2.Request(url)		
> f = urllib2.urlopen(req)
> 
> # POST Receipts data with json data
> data = <json data...>
> req = urllib2.Request(url, data=data, headers={'Content-Type': 'application/json'}) 	
> f = urllib2.urlopen(req)
> 
> When I run the same requests against my nginx installation the GET function works just fine.   However, the POST request returns a HTTP Error 400: BAD REQUEST error.  
> 
> It is exactly the same code and I know this is working with the django runserver.  It must be some configuration in Nginx that is regecting the POST request.  I get that error for any Content-Type.  
> 
> I have looked everywhere and I cannot find anything about allowing this post request.   I hope someone here can help nudge me in the right direction.  My Nginx config file settings is below.  I have modified many variables like buffer sizes and body sizes and it is not that.  The data I'm posting is less than 8k.  
> 
> Any ideas?
> 
> ----------------------------
> 
> server {
>        include mime.types;
>        default_type application/json;
> 
>        client_body_buffer_size 1200K;
>        client_header_buffer_size 1200K;
>        client_max_body_size 2m;
>        large_client_header_buffers 2 1200K;
> 
>        listen   80; ## listen for ipv4; this line is default and implied
>        #listen   [::]:80 default ipv6only=on; ## listen for ipv6
> 
> 
>        # Make site accessible from http://localhost/
>        server_name localhost;
> 
>        # Log information
>        access_log /home/henry/logs/access.log;
>        error_log /home/henry/logs/error.log;
> 
>        root /home/henry/webservice/public;
> 
>        location /site_media
>        {
>                root /path/to/yoursite.com/public;
>        }
> 
>        location / {
>                        # host and port to fastcgi server
>                        fastcgi_pass 127.0.0.1:8081;
>                        fastcgi_param PATH_INFO $fastcgi_script_name;
>                        fastcgi_param REQUEST_METHOD $request_method;
>                        fastcgi_param QUERY_STRING $query_string;
>                        fastcgi_param SERVER_NAME $server_name;
>                        fastcgi_param SERVER_PORT $server_port;
>                        fastcgi_param SERVER_PROTOCOL $server_protocol;
>                        fastcgi_param CONTENT_TYPE $content_type;
>                        fastcgi_param CONTENT_LENGTH $content_length;
>                        fastcgi_pass_header Authorization;
>                        fastcgi_intercept_errors off;
>        }
> 
>        location /doc {
>                root /usr/share;
>                autoindex on;
>                allow 127.0.0.1;
>                deny all;
>        }
> 
>        location /images {
>                root /usr/share;
>                autoindex off;
>        }
> 
> }
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120118/f389fc00/attachment-0001.html>

From robsonpeixoto at gmail.com  Wed Jan 18 17:40:12 2012
From: robsonpeixoto at gmail.com (Robson Roberto Souza Peixoto)
Date: Wed, 18 Jan 2012 15:40:12 -0200
Subject: Proxy with URI and params
In-Reply-To: <877h0qwl06.wl%appa@perusio.net>
References: <CAEA_=xJBsvzGj8tw7w0=vP097iqjGW8Mes1C8LGk=1c3LxeFzw@mail.gmail.com>
	<87fwffwgy1.wl%appa@perusio.net>
	<CAEA_=xJcx-y5HCkO3GCiD99gxVv0MnyePd84OzHX2B95Y+Q4PA@mail.gmail.com>
	<87aa5nw471.wl%appa@perusio.net> <878vl7w357.wl%appa@perusio.net>
	<CAEA_=xJsq3uciaiaqm7pn9moxTx1aPJNvLGZidw4N6tEhttkQg@mail.gmail.com>
	<877h0qwl06.wl%appa@perusio.net>
Message-ID: <CAEA_=xKagPKgvwbJWYs8Ry6rokkhi26m_N5UoCf0ZE6jYZwfRQ@mail.gmail.com>

Thanks for the tips. This solutions is a little slow, so
I'm studding another way to solve it.

Thanks a lot.

On Tue, Jan 17, 2012 at 10:50 AM, Ant?nio P. P. Almeida <appa at perusio.net>wrote:

> On 17 Jan 2012 11h27 WET, robsonpeixoto at gmail.com wrote:
>
> > I'd like to navigate on http://www.proxy.com/delivery/.../.../.../
> > using my domain http://www.example.com/.
> >
> > So the user will navigate on www.example.com and always show on URL
> > the ' www.example.com/delivery/../../..'
> >
> > It's is possible?
>
> Yes. Add the proxy_redirect directive:
>
> proxy_redirect http://www.proxy.com/delivery/ http://www.example.com/;
>
> See: http://wiki.nginx.org/HttpProxyModule#proxy_redirect
>
> --- appa
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>



-- 
Robson Roberto Souza Peixoto
Robinho
Master in Computer Science, University of Campinas
Linux Counter #395633
IRC: robsonpeixoto
Twitter: http://twitter.com/rrspba
github: https://github.com/robsonpeixoto
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120118/9e269b58/attachment.html>

From henry.ward at secondsightco.com  Wed Jan 18 19:24:22 2012
From: henry.ward at secondsightco.com (Henry Ward)
Date: Wed, 18 Jan 2012 11:24:22 -0800
Subject: Ningx/Django Bad Request Error - Partial SOLUTION
In-Reply-To: <201201182154.35707.ne@vbart.ru>
References: <86A60AE8-FAFB-451F-BDB9-24EEF89B9264@secondsightco.com>
	<7C5A503B-B113-47C6-AB83-26726A9FDB50@secondsightco.com>
	<201201182154.35707.ne@vbart.ru>
Message-ID: <BC47A02A-C727-4C26-BEB8-A14E8B844030@secondsightco.com>

Hi Valentin,

Thanks for the ideas.  I did the curl as you suggested and that works.  You were right that it is something wrong with the POST request using Python urllib.  I tried both request and httplib with urlib and neither worked.  The only thing that worked was a command line curl.

I checked the error path and all the settings and I am still unable to get this error to diagnosis further.  I am getting other errors in the log but not this one. 

I've never seen this problem before with Apache or django run server and it there is not much in the google searches about this.  Though I can hack the client with command like curl for now this seems Nginx issue that will reappear.

Henry



On Jan 18, 2012, at 9:54 AM, Valentin V. Bartenev wrote:

> On Wednesday 18 January 2012 20:58:48 you wrote:
> [...]
>> Error Log
>> The error log shows nothing.  Literally.  The access_log shows the access
>> url and the 400 error.  But the error log does not.  I know that is hard
>> to believe its similar to what is said here:
>> 
> 
> Did you set a correct path to error log in main context (not that one in your 
> server section) and debug level? Did you configure and build nginx with debug?
> 
> [...]
>> This is what makes so irritatingly difficult to debug.  It is also I feel
>> it some type of permission error or setting that does not allow POST
>> requests.  I've also read here (https://calomel.org/nginx.html) that Nginx
>> by default only allows GET requests but I couldn't find anywhere where the
>> is verified nor how to enable that.
> 
> https://calomel.org/nginx.html
> ## Only allow GET and HEAD request methods. By default Nginx blocks
> ## all requests type other then GET and HEAD for static content.
> 
> If you read carefully, you will notice "for static content". That means files in 
> your filesystem, obviously, cuz they can't process POST.
> 
> Actually, I expect that you POST request is just broken. It's ok for primitive 
> django builtin http-server, but it isn't acceptable for nginx. 
> 
> Why don't you use simple curl for emulate request? Please, try curl.
> 
> Or, it may be a django error, cuz of misconfigured fastcgi-param.
> Why don't you use stock fastcgi-params config that distributed with nginx? 
> 
> [...]
>> I have also created application logging in Django and have "GOT HERE" log
>> statement in the first line of the view that handles the POST request.. 
>> That line never gets called so it appears Nginx is kicking out the request
>> before it ever passes to the django app.
> 
> No. It just means that your request didn't reach the view, that handles POST 
> requests. Probably, because of misconfigured fastcgi-param django receive wrong
> or incomplete params.
> 
> Anyway, please, investigate and fix your problem with debug log. There's no 
> magic.
> 
> wbr, Valentin V. Bartenev


From henry.ward at secondsightco.com  Wed Jan 18 20:12:36 2012
From: henry.ward at secondsightco.com (Henry Ward)
Date: Wed, 18 Jan 2012 12:12:36 -0800
Subject: Ningx/Django Bad Request Error - Full Solution
In-Reply-To: <201201182154.35707.ne@vbart.ru>
References: <86A60AE8-FAFB-451F-BDB9-24EEF89B9264@secondsightco.com>
	<7C5A503B-B113-47C6-AB83-26726A9FDB50@secondsightco.com>
	<201201182154.35707.ne@vbart.ru>
Message-ID: <5D7346F8-36F8-4EFC-9266-CE0B31FF0678@secondsightco.com>

Found the problem.  A percent symbol snuck into the json data and I was having the same problem as below:

http://www.ruby-forum.com/topic/2573586

Other servers are more forgiving but Nginx was choking on urllib2's handling of escaping the % symbol.  

Sorry for the long winded discussion.

Henry

On Jan 18, 2012, at 9:54 AM, Valentin V. Bartenev wrote:

> On Wednesday 18 January 2012 20:58:48 you wrote:
> [...]
>> Error Log
>> The error log shows nothing.  Literally.  The access_log shows the access
>> url and the 400 error.  But the error log does not.  I know that is hard
>> to believe its similar to what is said here:
>> 
> 
> Did you set a correct path to error log in main context (not that one in your 
> server section) and debug level? Did you configure and build nginx with debug?
> 
> [...]
>> This is what makes so irritatingly difficult to debug.  It is also I feel
>> it some type of permission error or setting that does not allow POST
>> requests.  I've also read here (https://calomel.org/nginx.html) that Nginx
>> by default only allows GET requests but I couldn't find anywhere where the
>> is verified nor how to enable that.
> 
> https://calomel.org/nginx.html
> ## Only allow GET and HEAD request methods. By default Nginx blocks
> ## all requests type other then GET and HEAD for static content.
> 
> If you read carefully, you will notice "for static content". That means files in 
> your filesystem, obviously, cuz they can't process POST.
> 
> Actually, I expect that you POST request is just broken. It's ok for primitive 
> django builtin http-server, but it isn't acceptable for nginx. 
> 
> Why don't you use simple curl for emulate request? Please, try curl.
> 
> Or, it may be a django error, cuz of misconfigured fastcgi-param.
> Why don't you use stock fastcgi-params config that distributed with nginx? 
> 
> [...]
>> I have also created application logging in Django and have "GOT HERE" log
>> statement in the first line of the view that handles the POST request.. 
>> That line never gets called so it appears Nginx is kicking out the request
>> before it ever passes to the django app.
> 
> No. It just means that your request didn't reach the view, that handles POST 
> requests. Probably, because of misconfigured fastcgi-param django receive wrong
> or incomplete params.
> 
> Anyway, please, investigate and fix your problem with debug log. There's no 
> magic.
> 
> wbr, Valentin V. Bartenev


From nginx-forum at nginx.us  Wed Jan 18 22:02:02 2012
From: nginx-forum at nginx.us (mpratt)
Date: Wed, 18 Jan 2012 17:02:02 -0500
Subject: regex rewrite problem
Message-ID: <26fbf942e61f7dedb1753f66245ff27e.NginxMailingListEnglish@forum.nginx.org>

Hello,

I am new to this forum. Glad it exists. Thanks for that. I have a
production website that uses nginx 0.8.53 reverse proxy in front of
apache. I actually had little problem setting that up, but this simple
regex is giving me a headache.

For some reason, I am unable to take a URI like
example.com/manufacturers/metlox/0?page=13 and change it to
example.com/manufacturers/m   It seems like the curly braces aren?t
working.

I?ve tried the following without results:

location ~* /manufacturers/ {
        rewrite "^/manufacturers/(.{1})/(.*)$"
/manufacturer/directory/$1 redirect;
        }
location ~* /manufacturers/ {
        rewrite "^/manufacturers/([a-z] {1})/(.*)$"
/manufacturer/directory/$1 redirect;
        }

location ~* /manufacturers/ {
        rewrite "^/manufacturers/([a-z]?)/(.*)$"
/manufacturer/directory/$1 redirect;
        }

I am fairly new to regex, so maybe this is something simple. What I want
to get is the first letter of first captured word.  I am able to pull
the entire word with (.*) for $1, but not the first letter only. 

Any help would be greatly appreciated!

Thanks in advance.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221410,221410#msg-221410


From saqib.rasul at gmail.com  Wed Jan 18 22:06:00 2012
From: saqib.rasul at gmail.com (Saqib Rasul)
Date: Wed, 18 Jan 2012 23:06:00 +0100
Subject: nginx keep using more and more memory
In-Reply-To: <201201181652.23775.ne@vbart.ru>
References: <8d53edb63ae48f2c1bb1b904bde571e0.Nginx@www.serverphorums.com>
	<CANRKtT0xV89uoOm_HDvRqQHvbUfkxQaDKrZi+o9f1YRnD6K5kw@mail.gmail.com>
	<CANRKtT21n+XnX1WDDFNaZfcqm2+j-kSFH3rz9gnBLdApGEihPw@mail.gmail.com>
	<201201181652.23775.ne@vbart.ru>
Message-ID: <CANRKtT0o0pnEJ_UtJkhgkCwAejh8CTt6WTnv=Bi+nDCdxVYg-A@mail.gmail.com>

OK, i tired 1.1.13 and the bug seems to have been fixed there. Thanks
again for the help.

Regards,

Saqib

On Wed, Jan 18, 2012 at 1:52 PM, Valentin V. Bartenev <ne at vbart.ru> wrote:
> On Wednesday 18 January 2012 16:33:22 Saqib Rasul wrote:
>> Also do you know or can give an estimate of when the 1.1 version will go
>> stable?
>>
>
> http://trac.nginx.org/nginx/milestone/1.1.17
>
> FYI, nginx "devel" vs. "stable" difference mainly is about API and behavior
> stability. Both branches are reliable enough to use in production.
>
> ?wbr, Valentin V. Bartenev
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From kolesen.a at gmail.com  Wed Jan 18 22:32:43 2012
From: kolesen.a at gmail.com (Alexander Kolesen)
Date: Thu, 19 Jan 2012 01:32:43 +0300
Subject: regex rewrite problem
In-Reply-To: <26fbf942e61f7dedb1753f66245ff27e.NginxMailingListEnglish@forum.nginx.org>
References: <26fbf942e61f7dedb1753f66245ff27e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120118223242.GB22360@macbookpro.wargaming.net>

> Hello,
> 
> I am new to this forum. Glad it exists. Thanks for that. I have a
> production website that uses nginx 0.8.53 reverse proxy in front of
> apache. I actually had little problem setting that up, but this simple
> regex is giving me a headache.
> 
> For some reason, I am unable to take a URI like
> example.com/manufacturers/metlox/0?page=13 and change it to
> example.com/manufacturers/m   It seems like the curly braces aren?t
> working.
> 
> I?ve tried the following without results:
> 
> location ~* /manufacturers/ {
>         rewrite "^/manufacturers/(.{1})/(.*)$"
> /manufacturer/directory/$1 redirect;
>         }
> location ~* /manufacturers/ {
>         rewrite "^/manufacturers/([a-z] {1})/(.*)$"
> /manufacturer/directory/$1 redirect;
>         }
> 
> location ~* /manufacturers/ {
>         rewrite "^/manufacturers/([a-z]?)/(.*)$"
> /manufacturer/directory/$1 redirect;
>         }
> 
> I am fairly new to regex, so maybe this is something simple. What I want
> to get is the first letter of first captured word.  I am able to pull
> the entire word with (.*) for $1, but not the first letter only. 
> 
> Any help would be greatly appreciated!
> 
> Thanks in advance.
> 

Hello.

Try the following:

location /manufacturers/ {
    location ~ ^/manufacturers/(.) {
        return 301 /manufacturers/$1;
    }
}


From nginx-forum at nginx.us  Thu Jan 19 00:31:33 2012
From: nginx-forum at nginx.us (mpratt)
Date: Wed, 18 Jan 2012 19:31:33 -0500
Subject: regex rewrite problem
In-Reply-To: <26fbf942e61f7dedb1753f66245ff27e.NginxMailingListEnglish@forum.nginx.org>
References: <26fbf942e61f7dedb1753f66245ff27e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <52c873b5a419c3b8ff00c5a33de592b4.NginxMailingListEnglish@forum.nginx.org>

Brilliant! It works perfectly. Do you have any idea why the curly braces
didn't work and also where I can learn when to use a location in a
location? Thank you so much!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221410,221414#msg-221414


From nginx-forum at nginx.us  Thu Jan 19 01:43:08 2012
From: nginx-forum at nginx.us (Neddy)
Date: Wed, 18 Jan 2012 20:43:08 -0500
Subject: Reverse proxy for multiple websites
In-Reply-To: <9bfec7d5cd6030312300a3b9f019d1d1.NginxMailingListEnglish@forum.nginx.org>
References: <20120102235905.GD67687@mdounin.ru>
	<9bfec7d5cd6030312300a3b9f019d1d1.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <7689fb383d0ae875b4fe4c9eff6afcbd.NginxMailingListEnglish@forum.nginx.org>

dev Wrote:
-------------------------------------------------------
> hi,
> Can u show ur configuration?
> 
> Dev


Hi Dev,

I've followed this document
http://nginx.org/en/docs/http/request_processing.html

And I think my problem was missing deffinitions for "www.domain.tld" in
server_name; after edit it to

server_name domain.tld www.domain.tld; 

the problem solved.

Thanks for your attention.
~Neddy

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220736,221415#msg-221415


From appa at perusio.net  Thu Jan 19 01:47:44 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 19 Jan 2012 01:47:44 +0000
Subject: regex rewrite problem
In-Reply-To: <52c873b5a419c3b8ff00c5a33de592b4.NginxMailingListEnglish@forum.nginx.org>
References: <26fbf942e61f7dedb1753f66245ff27e.NginxMailingListEnglish@forum.nginx.org>
	<52c873b5a419c3b8ff00c5a33de592b4.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <8739bclayn.wl%appa@perusio.net>

On 19 Jan 2012 00h31 WET, nginx-forum at nginx.us wrote:

> Brilliant! It works perfectly. Do you have any idea why the curly
> braces didn't work and also where I can learn when to use a location
> in a location? Thank you so much!

Regexes with curly braces have to be quoted.

From the wiki: http://wiki.nginx.org/HttpRewriteModule#rewrite

     Note: for curly braces ('{' and '}'), as they are used both in
     regexes and for block control, to avoid conflicts, regexes with
     curly braces are to be enclosed with double quotes (or single
     quotes). For example, to rewrite URLs like

     rewrite  "/photos/([0-9] {2})([0-9] {2})([0-9] {2})" /path/to/photos/$1/$1$2/$1$2$3.png;

--- appa


From xb.zhou at qq.com  Thu Jan 19 02:45:52 2012
From: xb.zhou at qq.com (=?ISO-8859-1?B?WkhPVSBYaWFvYm8=?=)
Date: Thu, 19 Jan 2012 10:45:52 +0800
Subject: About nginx module developing
Message-ID: <tencent_5F1C554915774A504A1418BF@qq.com>

thanks, but I don't think SSI can satisfy my need. Because I want the RPS to be up to the limit of nginx it self.
My demand is easy and much like a http-push server:
1, I maintain 3 bitmaps
2, the clients request every 60 seconds
3, if the bit of the 3 bitmaps for this client is set, than forward to the upstream, or just response 200.

------------------
Sincerely yours
                         ZHOU Xiaobo
 
 
 
------------------ Original ------------------
From:  "Andrey N. Oktyabrski"<ano at bestmx.ru>;
Date:  Mon, Jan 16, 2012 06:19 PM
To:  "nginx"<nginx at nginx.org>;
Subject:  Re: About nginx module developing
 
On 15.01.12 19:31, ZHOU Xiaobo wrote:
> I am new to nginx module developing.
> I want to implement a module to deal with this demand:
>
> 1, nginx receives a request from client
> 2, nginx sends two internal requests(http or protobuf etc.) to 2 backend servers
> 3, after the responses of the two internal requests received nginx makes the response and sends to the client
>
> I am not sure that whether nginx can do this.
Maybe SSI is what you want?

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

<br>

From xb.zhou at qq.com  Thu Jan 19 03:22:12 2012
From: xb.zhou at qq.com (=?ISO-8859-1?B?WkhPVSBYaWFvYm8=?=)
Date: Thu, 19 Jan 2012 11:22:12 +0800
Subject: About nginx module developing
Message-ID: <tencent_0D29E5510F21E56436060DF1@qq.com>

And the 3 bitmap are provided through network, that is, 3 backend servers provide service
to test the bits in the bitmap.
So I want nginx to visit the 3 backend servers through module but not cgi-like things because of the 
high RPS from clients.

------------------
Sincerely yours
                         ZHOU Xiaobo
 
 
 
------------------ Original ------------------
From:  "xb.zhou"<xb.zhou at qq.com>;
Date:  Thu, Jan 19, 2012 10:45 AM
To:  "nginx"<nginx at nginx.org>;
Subject:  Re: About nginx module developing
 
thanks, but I don't think SSI can satisfy my need. Because I want the RPS to be up to the limit of nginx it self.
My demand is easy and much like a http-push server:
1, I maintain 3 bitmaps
2, the clients request every 60 seconds
3, if the bit of the 3 bitmaps for this client is set, than forward to the upstream, or just response 200.

------------------
Sincerely yours
                         ZHOU Xiaobo
 
 
 
------------------ Original ------------------
From:  "Andrey N. Oktyabrski"<ano at bestmx.ru>;
Date:  Mon, Jan 16, 2012 06:19 PM
To:  "nginx"<nginx at nginx.org>;
Subject:  Re: About nginx module developing
 
On 15.01.12 19:31, ZHOU Xiaobo wrote:
> I am new to nginx module developing.
> I want to implement a module to deal with this demand:
>
> 1, nginx receives a request from client
> 2, nginx sends two internal requests(http or protobuf etc.) to 2 backend servers
> 3, after the responses of the two internal requests received nginx makes the response and sends to the client
>
> I am not sure that whether nginx can do this.
Maybe SSI is what you want?

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

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

<br>

From nginx-forum at nginx.us  Thu Jan 19 03:46:25 2012
From: nginx-forum at nginx.us (mpratt)
Date: Wed, 18 Jan 2012 22:46:25 -0500
Subject: regex rewrite problem
In-Reply-To: <8739bclayn.wl%appa@perusio.net>
References: <8739bclayn.wl%appa@perusio.net>
Message-ID: <f5daa286d96d2ae6323ea1cc4d85d8cc.NginxMailingListEnglish@forum.nginx.org>

Thanks for your reply. I had previously read that section and enclosed
them in quotes and it didn't work:
for example:

> location ~* /manufacturers/ {
> rewrite "^/manufacturers/(.{1})/(.*)$" /manufacturer/directory/$1
redirect;
> }

doesn't work. That;'s why I was so stumped. Thought I had done
everything according to that section...

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221410,221420#msg-221420


From edho at myconan.net  Thu Jan 19 03:54:41 2012
From: edho at myconan.net (Edho Arief)
Date: Thu, 19 Jan 2012 10:54:41 +0700
Subject: regex rewrite problem
In-Reply-To: <f5daa286d96d2ae6323ea1cc4d85d8cc.NginxMailingListEnglish@forum.nginx.org>
References: <8739bclayn.wl%appa@perusio.net>
	<f5daa286d96d2ae6323ea1cc4d85d8cc.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CABTjkKkPFVRON70gfdX2d3D4EfQhUppiZF8aoHbSU3SoARz9dQ@mail.gmail.com>

On Thu, Jan 19, 2012 at 10:46 AM, mpratt <nginx-forum at nginx.us> wrote:
> Thanks for your reply. I had previously read that section and enclosed
> them in quotes and it didn't work:
> for example:
>
>> location ~* /manufacturers/ {
>> rewrite "^/manufacturers/(.{1})/(.*)$" /manufacturer/directory/$1
> redirect;
>> }
>
> doesn't work. That;'s why I was so stumped. Thought I had done
> everything according to that section...
>

Of course it won't. The regex above will only match url like this:

/manufacturers/a/something

and not:

/manufacturers/asdfgh/something

Because /(.{1})/ will only match (and capture) one character between
two slashes (e.g. /u/, /b/) while /(.).*/ will match (and capture)
first character of word between two slashes (e.g. /unix/, /books/).

-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From xujiecq at 163.com  Wed Jan 18 05:19:20 2012
From: xujiecq at 163.com (xujiecq)
Date: Wed, 18 Jan 2012 13:19:20 +0800
Subject: a question
Message-ID: <201201181319193598993@163.com>

I want to know. 
In source code?#define NGX_MAX_ALLOC_FROM_POOL (ngx_pagesize - 1)
Why decrement?




xujiecq
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120118/70c62449/attachment.html>

From nginx-forum at nginx.us  Thu Jan 19 07:06:32 2012
From: nginx-forum at nginx.us (sandeeppatel)
Date: Thu, 19 Jan 2012 02:06:32 -0500
Subject: Cross Compiling nginx-1.0.11
Message-ID: <f6fa8100742737c681eb5897d9df531a.NginxMailingListEnglish@forum.nginx.org>

Hi ,
I want to know how to cross compile nginx-1.0.11.

I have already tried :  --with-cc option
but it is giving me an error : checking for C compiler ... found but is
not working

                                        ./configure: error: C compiler
arm-linux-gcc is not found
                                        
                                 --with-cc-opt is also not working.
I tried to change compiler manually too(by changing a Makefile).

I have already install arm-linux-gcc and its working fine with
openssl(It successfully cross compiled openssl. so problem is not with
arm-linux compiler).

Any help will be highly appreciated .. 
Thanks in advance. :)

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221430,221430#msg-221430


From m.pratt1 at verizon.net  Wed Jan 18 16:58:27 2012
From: m.pratt1 at verizon.net (mpratt)
Date: Wed, 18 Jan 2012 11:58:27 -0500
Subject: regex issue
Message-ID: <000001ccd602$6636fc60$32a4f520$@verizon.net>

For some reason, I am unable to take a URI like
example.com/manufacturers/metlox/0?page=13 and change it to
example.com/manufacturers/m

It seems like the curly braces aren't working. I've tried the following
without results:

 

location ~* /manufacturers/ {

        rewrite "^/manufacturers/(.{1})/(.*)$" /manufacturer/directory/$1
redirect;

        }

location ~* /manufacturers/ {

        rewrite "^/manufacturers/([a-z] {1})/(.*)$"
/manufacturer/directory/$1 redirect;

        }

 

location ~* /manufacturers/ {

        rewrite "^/manufacturers/([a-z]?)/(.*)$" /manufacturer/directory/$1
redirect;

        }

 

I am new to regex, so maybe this is something simple. What I want to get is
the first letter of first captured word.  What I can do is pull the entire
word with (.*), but not the first letter only. 

Any help would be appreciated.

 

Thanks in advance. 

 

Using nginx 0.8.53 reverse proxy in front of apache

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120118/b59f461d/attachment.html>

From edho at myconan.net  Thu Jan 19 07:17:10 2012
From: edho at myconan.net (Edho Arief)
Date: Thu, 19 Jan 2012 14:17:10 +0700
Subject: regex issue
In-Reply-To: <000001ccd602$6636fc60$32a4f520$@verizon.net>
References: <000001ccd602$6636fc60$32a4f520$@verizon.net>
Message-ID: <CABTjkKkv=psHfw2wyh-ZENyfx1BjaEtaqcBdgUwkBTO52idFrQ@mail.gmail.com>

On Wed, Jan 18, 2012 at 11:58 PM, mpratt <m.pratt1 at verizon.net> wrote:
> For some reason, I am unable to take a URI like
> example.com/manufacturers/metlox/0?page=13 and change it to
> example.com/manufacturers/m
>
> It seems like the curly braces aren?t working. I?ve tried the following
> without results:
>



/(.{1})/
This will match and capture one character between slashes
Examples:
/a/ -> $1: a
/asd/ -> no match (three characters)

/([a-z]*)/
This one will match and capture one or less letter between slashes
Examples:
/a/ -> $1: a
/asd/ -> no match

/([a-z]?)/
This one is equivalent to /([a-z]*)/

/([a-z] ?)/
This one will match and capture one letter followed by space between slashes
Examples:
/a/ -> no match
/a / -> $1: " a"
/asd/ -> no match

/(.).*/
This one will match one or more characters between slashes and capture
the first character
/a/ -> $1: a
/a / -> $1: a
/asd/ -> $1: a


-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From edho at myconan.net  Thu Jan 19 07:18:40 2012
From: edho at myconan.net (Edho Arief)
Date: Thu, 19 Jan 2012 14:18:40 +0700
Subject: regex issue
In-Reply-To: <CABTjkKkv=psHfw2wyh-ZENyfx1BjaEtaqcBdgUwkBTO52idFrQ@mail.gmail.com>
References: <000001ccd602$6636fc60$32a4f520$@verizon.net>
	<CABTjkKkv=psHfw2wyh-ZENyfx1BjaEtaqcBdgUwkBTO52idFrQ@mail.gmail.com>
Message-ID: <CABTjkKnDPBdtcWGNboHP4yfAGgBLHU001WPP0H0+jBd5NSe6dg@mail.gmail.com>

Sorry, I made mistake
>
> /([a-z] ?)/
> This one will match and capture one letter optionally followed by space between slashes
> Examples:
> /a/ -> $1: a
> /a / -> $1: "a "
> /asd/ -> no match
>


-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From nginx-forum at nginx.us  Thu Jan 19 08:21:08 2012
From: nginx-forum at nginx.us (ilin)
Date: Thu, 19 Jan 2012 03:21:08 -0500
Subject: Static index file in Nginx+Apache
Message-ID: <568a64b04236d347c01558df9511a7e7.NginxMailingListEnglish@forum.nginx.org>

location / {
	index index.htm;
	proxy_pass   http://127.0.0.1:8080;
	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;
	}

location ~* \.(jpg|jpeg|gif|png|ico|css|bmp|js|swf|html|htm)$ {
	root /home/site/public_html/;
}

I need process the request "GET /" via Nginx (as static file).
Added directive "index" has no effect, the request passed to backend.
What's wrong?

Thanks.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221434,221434#msg-221434


From edho at myconan.net  Thu Jan 19 08:31:54 2012
From: edho at myconan.net (Edho Arief)
Date: Thu, 19 Jan 2012 15:31:54 +0700
Subject: Static index file in Nginx+Apache
In-Reply-To: <568a64b04236d347c01558df9511a7e7.NginxMailingListEnglish@forum.nginx.org>
References: <568a64b04236d347c01558df9511a7e7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CABTjkKkuaebhhMx4MpNte1JTRAgOg0kGShDnHBAtDxq6H=JUXQ@mail.gmail.com>

On Thu, Jan 19, 2012 at 3:21 PM, ilin <nginx-forum at nginx.us> wrote:
> location / {
> ? ? ? ?index index.htm;
> ? ? ? ?proxy_pass ? http://127.0.0.1:8080;
> ? ? ? ?proxy_set_header ? X-Real-IP $remote_addr;
> ? ? ? ?proxy_set_header ? Host $http_host;
> ? ? ? ?proxy_set_header ? X-Forwarded-For $proxy_add_x_forwarded_for;
> ? ? ? ?}
>
> location ~* \.(jpg|jpeg|gif|png|ico|css|bmp|js|swf|html|htm)$ {
> ? ? ? ?root /home/site/public_html/;
> }
>
> I need process the request "GET /" via Nginx (as static file).
> Added directive "index" has no effect, the request passed to backend.
> What's wrong?
>

since "/" doesn't match the latter part, just add this one location block:

location = / { }

-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From albert at nn.iij4u.or.jp  Thu Jan 19 09:37:40 2012
From: albert at nn.iij4u.or.jp (shin fukuda)
Date: Thu, 19 Jan 2012 18:37:40 +0900
Subject: cache object list
Message-ID: <20120119183740.4b7dae49b0559fad7c0a2328@nn.iij4u.or.jp>

Hi.

How do i get Nginx's cache object list ?



-- 
shin fukuda <albert at nn.iij4u.or.jp>


From nginx-forum at nginx.us  Thu Jan 19 09:40:51 2012
From: nginx-forum at nginx.us (ilin)
Date: Thu, 19 Jan 2012 04:40:51 -0500
Subject: Static index file in Nginx+Apache
In-Reply-To: <CABTjkKkuaebhhMx4MpNte1JTRAgOg0kGShDnHBAtDxq6H=JUXQ@mail.gmail.com>
References: <CABTjkKkuaebhhMx4MpNte1JTRAgOg0kGShDnHBAtDxq6H=JUXQ@mail.gmail.com>
Message-ID: <5f26ffc0f310beebf98d2bf03c36a9ac.NginxMailingListEnglish@forum.nginx.org>

Nginx responded 404

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221434,221439#msg-221439


From edho at myconan.net  Thu Jan 19 10:50:36 2012
From: edho at myconan.net (Edho Arief)
Date: Thu, 19 Jan 2012 17:50:36 +0700
Subject: Static index file in Nginx+Apache
In-Reply-To: <5f26ffc0f310beebf98d2bf03c36a9ac.NginxMailingListEnglish@forum.nginx.org>
References: <CABTjkKkuaebhhMx4MpNte1JTRAgOg0kGShDnHBAtDxq6H=JUXQ@mail.gmail.com>
	<5f26ffc0f310beebf98d2bf03c36a9ac.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CABTjkKkFUdoUrgYFmEKZjkTiYdRJ0EM7WPTy8saaE0RDKEH39A@mail.gmail.com>

On Thu, Jan 19, 2012 at 4:40 PM, ilin <nginx-forum at nginx.us> wrote:
> Nginx responded 404
>

http://wiki.nginx.org/Pitfalls#Root_inside_Location_Block


-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From mdounin at mdounin.ru  Thu Jan 19 10:58:42 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 19 Jan 2012 14:58:42 +0400
Subject: Cross Compiling nginx-1.0.11
In-Reply-To: <f6fa8100742737c681eb5897d9df531a.NginxMailingListEnglish@forum.nginx.org>
References: <f6fa8100742737c681eb5897d9df531a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120119105842.GG67687@mdounin.ru>

Hello!

On Thu, Jan 19, 2012 at 02:06:32AM -0500, sandeeppatel wrote:

> Hi ,
> I want to know how to cross compile nginx-1.0.11.
> 
> I have already tried :  --with-cc option
> but it is giving me an error : checking for C compiler ... found but is
> not working
> 
>                                         ./configure: error: C compiler
> arm-linux-gcc is not found
>                                         
>                                  --with-cc-opt is also not working.
> I tried to change compiler manually too(by changing a Makefile).
> 
> I have already install arm-linux-gcc and its working fine with
> openssl(It successfully cross compiled openssl. so problem is not with
> arm-linux compiler).
> 
> Any help will be highly appreciated .. 
> Thanks in advance. :)

Cross-compilation isn't currently supported by nginx.

Maxim Dounin


From nginx-forum at nginx.us  Thu Jan 19 12:24:29 2012
From: nginx-forum at nginx.us (bigplum)
Date: Thu, 19 Jan 2012 07:24:29 -0500
Subject: how to make a named location in subrequest
Message-ID: <e0fcce27e1f88f8bb28138d2d58cb611.NginxMailingListEnglish@forum.nginx.org>

Hi,

Previously, I wrote a module to make a subrequest to backend. nginx.conf
like that:

location /file {
    nphase_uri /dummy;
    nphase_set_uri_var $np_uri http://backend;
}

location /dummy {
    proxy_pass $np_uri;
}

The module could send original URI "http://localhost/file/abcd.txt" to
backend to process.
And the module could get response from backend and decide to continue or
refuse.

But now This fix disabled sending original URI to backend.

    *) Bugfix: a "proxy_pass" directive without URI part might use
original
       request after redirection with the "try_files" directive.

If we need to send same request to backend, we must use named location.
But it
seems that named location is not support in subrequest. The named
location only
processed in try_files phase.

Following configuration would not work, for the subrequest send a
request "GET /dummy"
rather than original URI.

location /dummy {
    try_files /dummy @named;
}
location @named {
    proxy_pass $np_uri;
}

So how do I make it work?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221447,221447#msg-221447


From nginx-forum at nginx.us  Thu Jan 19 13:30:20 2012
From: nginx-forum at nginx.us (mpratt)
Date: Thu, 19 Jan 2012 08:30:20 -0500
Subject: regex rewrite problem
In-Reply-To: <26fbf942e61f7dedb1753f66245ff27e.NginxMailingListEnglish@forum.nginx.org>
References: <26fbf942e61f7dedb1753f66245ff27e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <acc5e9dfac564cb41c672ca495f4f3dd.NginxMailingListEnglish@forum.nginx.org>

Thank you for this explanation!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221410,221449#msg-221449


From nginx-forum at nginx.us  Thu Jan 19 13:35:41 2012
From: nginx-forum at nginx.us (mpratt)
Date: Thu, 19 Jan 2012 08:35:41 -0500
Subject: regex issue
In-Reply-To: <000001ccd602$6636fc60$32a4f520$@verizon.net>
References: <000001ccd602$6636fc60$32a4f520$@verizon.net>
Message-ID: <da9d5e94aee6a12c46c3bd141f801538.NginxMailingListEnglish@forum.nginx.org>

Thank you so much. This explanation really helps. 

Sorry for the nearly duplicate post. This one came through the mailing
list delayed before I discovered this forum.

Thanks again!

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221431,221450#msg-221450


From ionathan at gmail.com  Thu Jan 19 14:44:18 2012
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Thu, 19 Jan 2012 11:44:18 -0300
Subject: proxy_pass by the accept header?
Message-ID: <CAO6fm-S9vN5Ep6MwOAfgB3BOxrHGvdauGbDGnM8sA0z3nuEAHg@mail.gmail.com>

Hi all!
I want to proxy_pass depending on the value of the accept header. I have
lots of locations and I wanted to add in a single place a validation that
if accept header is 'text/html" it should proxy_pass to some specific
upstream. Else, just go ahead and try to match a location.

Is that possible?

Thanks!

Jonathan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120119/efe105b7/attachment.html>

From appa at perusio.net  Thu Jan 19 15:22:33 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Thu, 19 Jan 2012 15:22:33 +0000
Subject: proxy_pass by the accept header?
In-Reply-To: <CAO6fm-S9vN5Ep6MwOAfgB3BOxrHGvdauGbDGnM8sA0z3nuEAHg@mail.gmail.com>
References: <CAO6fm-S9vN5Ep6MwOAfgB3BOxrHGvdauGbDGnM8sA0z3nuEAHg@mail.gmail.com>
Message-ID: <87zkdjk98m.wl%appa@perusio.net>

On 19 Jan 2012 14h44 WET, ionathan at gmail.com wrote:

> Hi all!  I want to proxy_pass depending on the value of the accept
> header. I have lots of locations and I wanted to add in a single
> place a validation that if accept header is 'text/html" it should
> proxy_pass to some specific upstream. Else, just go ahead and try to
> match a location.

> Is that possible?

Perhaps something like:

# http level

map $http_accept $my_upstream {
   default misc_upstream;
   ~text/html html_upstream;
   
}

upstream misc_upstream {
    server 127.0.0.1:8080;     
}

upstream html_upstream {
    server 127.0.0.1:8088;     
}

# server level

location / {
    #... your stuff
    proxy_pass http://$my_upstream;
}

--- appa


From ionathan at gmail.com  Thu Jan 19 15:39:34 2012
From: ionathan at gmail.com (Jonathan Leibiusky)
Date: Thu, 19 Jan 2012 12:39:34 -0300
Subject: proxy_pass by the accept header?
In-Reply-To: <87zkdjk98m.wl%appa@perusio.net>
References: <CAO6fm-S9vN5Ep6MwOAfgB3BOxrHGvdauGbDGnM8sA0z3nuEAHg@mail.gmail.com>
	<87zkdjk98m.wl%appa@perusio.net>
Message-ID: <CAO6fm-RejamCsBJ4eC2DUE-5PGk4bD8+O4b=aK504P1YvXnpCw@mail.gmail.com>

this is a great solution!
thanks so much!

On Thu, Jan 19, 2012 at 12:22 PM, Ant?nio P. P. Almeida <appa at perusio.net>wrote:

> On 19 Jan 2012 14h44 WET, ionathan at gmail.com wrote:
>
> > Hi all!  I want to proxy_pass depending on the value of the accept
> > header. I have lots of locations and I wanted to add in a single
> > place a validation that if accept header is 'text/html" it should
> > proxy_pass to some specific upstream. Else, just go ahead and try to
> > match a location.
>
> > Is that possible?
>
> Perhaps something like:
>
> # http level
>
> map $http_accept $my_upstream {
>   default misc_upstream;
>   ~text/html html_upstream;
>
> }
>
> upstream misc_upstream {
>    server 127.0.0.1:8080;
> }
>
> upstream html_upstream {
>    server 127.0.0.1:8088;
> }
>
> # server level
>
> location / {
>    #... your stuff
>    proxy_pass http://$my_upstream;
> }
>
> --- appa
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120119/bbc7c10e/attachment.html>

From mdounin at mdounin.ru  Thu Jan 19 16:06:26 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 19 Jan 2012 20:06:26 +0400
Subject: how to make a named location in subrequest
In-Reply-To: <e0fcce27e1f88f8bb28138d2d58cb611.NginxMailingListEnglish@forum.nginx.org>
References: <e0fcce27e1f88f8bb28138d2d58cb611.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120119160626.GQ67687@mdounin.ru>

Hello!

On Thu, Jan 19, 2012 at 07:24:29AM -0500, bigplum wrote:

> Hi,
> 
> Previously, I wrote a module to make a subrequest to backend. nginx.conf
> like that:
> 
> location /file {
>     nphase_uri /dummy;
>     nphase_set_uri_var $np_uri http://backend;
> }
> 
> location /dummy {
>     proxy_pass $np_uri;
> }
> 
> The module could send original URI "http://localhost/file/abcd.txt" to
> backend to process.
> And the module could get response from backend and decide to continue or
> refuse.
> 
> But now This fix disabled sending original URI to backend.
> 
>     *) Bugfix: a "proxy_pass" directive without URI part might use
> original
>        request after redirection with the "try_files" directive.
> 
> If we need to send same request to backend, we must use named location.
> But it
> seems that named location is not support in subrequest. The named
> location only
> processed in try_files phase.
> 
> Following configuration would not work, for the subrequest send a
> request "GET /dummy"
> rather than original URI.
> 
> location /dummy {
>     try_files /dummy @named;
> }
> location @named {
>     proxy_pass $np_uri;
> }
> 
> So how do I make it work?

You may explicitly specify URI you want to use in a request to 
backend, i.e.

    location /dummy {
        proxy_pass $np_uri$request_uri;
    }

Maxim Dounin


From siefke_listen at web.de  Fri Jan 20 00:55:47 2012
From: siefke_listen at web.de (siefke_listen at web.de)
Date: Fri, 20 Jan 2012 01:55:47 +0100
Subject: Nginx Vhost
Message-ID: <20120120015547.fb552ea4.siefke_listen@web.de>

Hello, 

i have installed Nginx 1.0.11 on FreeBSD 9. I have configured a Catch All
Host and my Domain. But my Domain Nginx want not accept. He give only
the Catch All Domain back. 

<config>
user  www;
worker_processes  1;
error_log  logs/error.log;

events {
    worker_connections  1024;
}


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

log_format  main  '$remote_addr - $remote_user [$time_local] "$request" '
                  '$status $body_bytes_sent "$http_referer" '
                  '"$http_user_agent" "$http_x_forwarded_for"';

access_log  /var/log/www/access.log  main;

sendfile        on;
keepalive_timeout  65;

server {
listen       80;
server_name  _;
access_log   logs/access.log  main;

location / {
root   /usr/local/www/nginx;
index  index.html index.htm;
}

error_page  404              /404.html;
error_page   500 502 503 504  /50x.html;
location = /50x.html {
root   /usr/local/www/nginx-dist;
}
}

#silviosiefke.de
server {
listen	     80;
server_name  silviosiefke.de;
access_log   /usr/local/www/vhost/silviosiefke_de/log/access.log main;
error_log    /usr/local/www/vhost/silviosiefke_de/log/error.log;


location / {
root   /usr/local/www/vhost/silviosiefke_de/html;
index  index.html index.htm index.php;
}

error_page   500 502 503 504  /50x.html;

location = /50x.html {
root   /usr/local/www/nginx-dist;
}		
	
#location /article-(\d+)\.html$ {
#try_files $uri $uri/ /blogdetail.html?id=$1;
#}
	
# pass the PHP scripts to FastCGI 
#location ~ \.(php|htm|html)$ {
#fastcgi_pass   127.0.0.1:9000;
#fastcgi_index  index.php;
#fastcgi_param SCRIPT_FILENAME  /usr/local/www/vhost/silviosiefke_de/html
$fastcgi_script_name; 
#include /usr/local/etc/nginx/fastcgi_params; #}

#location ~ /\.ht {
#deny  all;
#}
}
}
</config>

When i give the ip nginx use root /usr/local/www/nginx and when i call
silviosiefke.de, nginx give back the root /usr/local/www/nginx. 

Where is the mistake? Thanks for help.


Regards
Silvio 


From nginx-forum at nginx.us  Fri Jan 20 08:04:30 2012
From: nginx-forum at nginx.us (bigplum)
Date: Fri, 20 Jan 2012 03:04:30 -0500
Subject: how to make a named location in subrequest
In-Reply-To: <e0fcce27e1f88f8bb28138d2d58cb611.NginxMailingListEnglish@forum.nginx.org>
References: <e0fcce27e1f88f8bb28138d2d58cb611.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <cd47b40dfa669b044eadf4be47c8cfca.NginxMailingListEnglish@forum.nginx.org>

Thanks to Maxim Dounin.

Finally I modify the module. Cat the URI string to $np_uri, make $np_uri
like that http://backend/filename.ext.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221447,221467#msg-221467


From nginx-forum at nginx.us  Fri Jan 20 08:23:24 2012
From: nginx-forum at nginx.us (bigplum)
Date: Fri, 20 Jan 2012 03:23:24 -0500
Subject: how to make a named location in subrequest
In-Reply-To: <e0fcce27e1f88f8bb28138d2d58cb611.NginxMailingListEnglish@forum.nginx.org>
References: <e0fcce27e1f88f8bb28138d2d58cb611.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <c505a061ad622d5f418c84913ba10498.NginxMailingListEnglish@forum.nginx.org>

The  r->unparsed_uri has been used to cat.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221447,221468#msg-221468


From lists at ruby-forum.com  Fri Jan 20 10:04:20 2012
From: lists at ruby-forum.com (kamalakar v.)
Date: Fri, 20 Jan 2012 11:04:20 +0100
Subject: Need to proxy pass and rewrite the url
Message-ID: <a9edc7bd53268a008ab87d98eb819cde@ruby-forum.com>

my conf :

 location / {
        proxy_pass           http://mydomain.com;
        proxy_redirect       off;
        proxy_set_header     Host $host;
    }

    location ~*/redirect {
        proxy_pass           http://some.domain.com/cp/$1;
    }


* http://mydomain.com/12121/redirect -->>
http://some.domain.com/cp/redirect

  * if the url contains redirect then i should rewrite the url ( need to
change the domain and append 'cp' to the url

But with above config not working giving me following error page
 "The page you are looking for is temporarily unavailable. Please try
again later."


Referred the below thread:
 http://www.ruby-forum.com/topic/144179
 This almost clear my problem but need to append cp to the url?

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


From nginx-forum at nginx.us  Fri Jan 20 10:54:01 2012
From: nginx-forum at nginx.us (supportnti)
Date: Fri, 20 Jan 2012 05:54:01 -0500
Subject: what is ngx_cycle_t represent
In-Reply-To: <20120110104336.GM67687@mdounin.ru>
References: <20120110104336.GM67687@mdounin.ru>
Message-ID: <f7a490aba44f26dd82e8bd4bf958414d.NginxMailingListEnglish@forum.nginx.org>

Thanks Obi-Wan!

I am going thru the source as we speak and would appreciate a helping
hand!

I can across "NGINX" environment variable in the function
"ngx_add_inherited_sockets". What does the NGINX variable represent? And
what does the function do?

Appreciate your response.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220954,221473#msg-221473


From nginx-forum at nginx.us  Fri Jan 20 12:14:02 2012
From: nginx-forum at nginx.us (DeineAgentur)
Date: Fri, 20 Jan 2012 07:14:02 -0500
Subject: htaccess to nginx
Message-ID: <ba22d7fe8b4cf81092a03227cd08b0b3.NginxMailingListEnglish@forum.nginx.org>

Hey guys!

i have following htaccess but now i need it for nginx

can some one convert it correct to nginx so that ther are no security
risks?



ErrorDocument 404 /404.html
AddDefaultCharset utf-8
AddCharset utf-8 .html .php .css
Options FollowSymLinks

AddHandler application/x-httpd-php .html
<FilesMatch "\.(htm|html|php)$">
AddDefaultCharset UTF-8
# php_value auto_prepend_file
/home/webpages/lima-city/deineagentur/html/DeineAgentur/zip_output_header.php
</FilesMatch>

RewriteEngine On
# RewriteCond %{HTTP_HOST} !^oceangame\.deineagentur\.com$ [NC]
# RewriteRule ^(.*)$ http://oceangame.deineagentur.com/$1 [L,R=301]

RewriteCond %{REQUEST_METHOD} ^(TRACE|TRACK|OPTIONS|HEAD)$ [NC]
RewriteRule ^.*$ - [L,F]


####################
# URI redirections #
####################
# Move to www.example.com (we don't wanna serve this page without the
subdomain www)
RewriteCond %{HTTP_HOST} !^oceangame.deineagentur.com [NC]
RewriteRule ^.*$ http://oceangame.deineagentur.com%{REQUEST_URI}
[L,NC,R=301]
# If so, then load language defined by cookie
RewriteCond %{REQUEST_URI} ^/$
RewriteCond %{HTTP_COOKIE} lang=([a-z]{2})
RewriteRule ^.*$ /%1 [L,NC,QSA,R=301]
# Add a trailing slash to the request uri if /de was requested without
one
RewriteCond %{REQUEST_URI} ^/(de|en|fr|es|it)$
RewriteRule ^.*$ /%1/ [L,NC,QSA,R=301]

####################################
# Request the appropriate language #
####################################
# German
RewriteCond %{REQUEST_URI} ^/de(/)?$
RewriteRule ^de/?.*$ /system.php?og__site=index
[L,NC,QSA,E=LANGNAME:de,E=SYSTEM:portal]
RewriteCond %{REQUEST_URI} ^/de/([a-z]+).html$
RewriteRule ^de/([a-z]+).html$ /system.php?og__site=$1
[L,NC,QSA,E=LANGNAME:de,E=SYSTEM:portal]
# English
RewriteCond %{REQUEST_URI} ^/en(/)?$
RewriteRule ^en/?.*$ /system.php?og__site=index
[L,NC,QSA,E=LANGNAME:en,E=SYSTEM:portal]
RewriteCond %{REQUEST_URI} ^/en/([a-z]+).html$
RewriteRule ^en/([a-z]+).html$ /system.php?og__site=$1
[L,NC,QSA,E=LANGNAME:en,E=SYSTEM:portal]
# French
RewriteCond %{REQUEST_URI} ^/fr(/)?$
RewriteRule ^fr/?.*$ /system.php?og__site=index
[L,NC,QSA,E=LANGNAME:fr,E=SYSTEM:portal]
RewriteCond %{REQUEST_URI} ^/fr/([a-z]+).html$
RewriteRule ^fr/([a-z]+).html$ /system.php?og__site=$1
[L,NC,QSA,E=LANGNAME:fr,E=SYSTEM:portal]
# Spain
RewriteCond %{REQUEST_URI} ^/es(/)?$
RewriteRule ^es/?.*$ /system.php?og__site=index
[L,NC,QSA,E=LANGNAME:es,E=SYSTEM:portal]
RewriteCond %{REQUEST_URI} ^/es/([a-z]+).html$
RewriteRule ^es/([a-z]+).html$ /system.php?og__site=$1
[L,NC,QSA,E=LANGNAME:es,E=SYSTEM:portal]
# Italian
RewriteCond %{REQUEST_URI} ^/it(/)?$
RewriteRule ^it/?.*$ /system.php?og__site=index
[L,NC,QSA,E=LANGNAME:it,E=SYSTEM:portal]
RewriteCond %{REQUEST_URI} ^/it/([a-z]+).html$
RewriteRule ^it/([a-z]+).html$ /system.php?og__site=$1
[L,NC,QSA,E=LANGNAME:it,E=SYSTEM:portal]


########################
# Serving normal files #
########################
# Request language specific sitemap.xml
RewriteCond %{REQUEST_URI} ^/(de|en|fr|es|it)/sitemap.xml$
RewriteRule ^(de|en|fr|es|it)/.*$ /sitemap_$1.xml [L,NC]
# Request normal files like images or style sheets
RewriteCond %{IS_SUBREQ} false
RewriteRule ^(?:de|en|fr|es|it)/(.*)$ /$1 [L,NC,QSA,R=301]
# Move to index.php if the requested file doesn't exist
RewriteCond %{REQUEST_FILENAME} !-f
RewriteCond %{REQUEST_FILENAME} !-d
RewriteRule ^.*$ http://oceangame.deineagentur.com [L,NC,QSA,R=301]

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221476,221476#msg-221476


From mdounin at mdounin.ru  Fri Jan 20 13:20:14 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Fri, 20 Jan 2012 17:20:14 +0400
Subject: what is ngx_cycle_t represent
In-Reply-To: <f7a490aba44f26dd82e8bd4bf958414d.NginxMailingListEnglish@forum.nginx.org>
References: <20120110104336.GM67687@mdounin.ru>
	<f7a490aba44f26dd82e8bd4bf958414d.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120120132014.GB67687@mdounin.ru>

Hello!

On Fri, Jan 20, 2012 at 05:54:01AM -0500, supportnti wrote:

> Thanks Obi-Wan!
> 
> I am going thru the source as we speak and would appreciate a helping
> hand!
> 
> I can across "NGINX" environment variable in the function
> "ngx_add_inherited_sockets". What does the NGINX variable represent? And
> what does the function do?

The "NGINX" environment variable is used to pass listening sockets 
from an old master process to a new one during binary upgrade 
procedure.  (Or, more strictly, it holds numbers of file 
descriptors.  Sockets remain opened across execve() call, but 
new process needs to know their numbers.)

The ngx_add_inherited_sockets() function parses the environment 
variable in question and adds sockets to the list of listening 
sockets.

Maxim Dounin


From appa at perusio.net  Fri Jan 20 14:24:45 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Fri, 20 Jan 2012 14:24:45 +0000
Subject: Need to proxy pass and rewrite the url
In-Reply-To: <a9edc7bd53268a008ab87d98eb819cde@ruby-forum.com>
References: <a9edc7bd53268a008ab87d98eb819cde@ruby-forum.com>
Message-ID: <87r4yujvte.wl%appa@perusio.net>

On 20 Jan 2012 10h04 WET, lists at ruby-forum.com wrote:

> my conf :
>
> location / {
> proxy_pass           http://mydomain.com;
> proxy_redirect       off;
> proxy_set_header     Host $host;
> }
>
> location ~*/redirect {
> proxy_pass           http://some.domain.com/cp/$1;
> }
>

This makes no sense. What do you want to capture in $1?

--- appa


From nginxyz at mail.ru  Fri Jan 20 21:36:15 2012
From: nginxyz at mail.ru (=?UTF-8?B?TWF4?=)
Date: Sat, 21 Jan 2012 01:36:15 +0400
Subject: Reading from a named pipe / fifo
Message-ID: <E1RoM87-0006bo-67.nginxyz-mail-ru@f297.mail.ru>


Hi,

how could I configure (or change) nginx to serve the output
from a named pipe / fifo (that was created with mknod or mkfifo)
as if it were a regular HTML file?

Let's say I have a fifo named fifo.html that gets input from an
application. This input will sometimes be read locally by another
application and sometimes remotely through nginx, but once the  
information is read, it should be considered consumed (no longer
available).       

Can the current version of nginx be configured to provide this 
feature?

Limiting the number of times a certain file can be served would
also be an acceptable solution.

If this isn't possible, could you point me in the right direction
so I could change the source code or write a module that would
allow nginx to serve output from named pipes / fifos?

Thank you, 
Max

From brian at akins.org  Fri Jan 20 22:21:10 2012
From: brian at akins.org (Brian Akins)
Date: Fri, 20 Jan 2012 17:21:10 -0500
Subject: Reading from a named pipe / fifo
In-Reply-To: <E1RoM87-0006bo-67.nginxyz-mail-ru@f297.mail.ru>
References: <E1RoM87-0006bo-67.nginxyz-mail-ru@f297.mail.ru>
Message-ID: <CAO3qTJXN7mROOz_O4p8iFV09KtUhmNPosRvz+wnS7gx80OiXhA@mail.gmail.com>

nginx can http proxy to a unix socket. (as well as fastcgi, uwsgi, etc). So
could the other app provide the data that way?  With a fifo, you'd need a
way to specify when to stop reading, so you may as well use a known
protocol.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120120/bb9cb4af/attachment.html>

From nginxyz at mail.ru  Sat Jan 21 16:28:15 2012
From: nginxyz at mail.ru (=?UTF-8?B?TWF4?=)
Date: Sat, 21 Jan 2012 20:28:15 +0400
Subject: Reading from a named pipe / fifo
In-Reply-To: <CAO3qTJXN7mROOz_O4p8iFV09KtUhmNPosRvzwnS7gx80OiXhA@mail.gmail.com>
References: <E1RoM87-0006bo-67.nginxyz-mail-ru@f297.mail.ru>
	<CAO3qTJXN7mROOz_O4p8iFV09KtUhmNPosRvzwnS7gx80OiXhA@mail.gmail.com>
Message-ID: <E1Rodnb-00028Q-L0.nginxyz-mail-ru@f167.mail.ru>

 21 ?????? 2012, 02:21 ?? Brian Akins : > nginx can http proxy to a unix socket. (as well as fastcgi, uwsgi, etc). So > could the other app provide the data that way? With a fifo, you'd need a > way to specify when to stop reading, so you may as well use a known > protocol. Unfortunately, the other app can provide the data only through a fifo and there's no easy way to change it due to licensing, so writing a fifo module for nginx would be the easier and cleaner solution. Igor and other developers, could you please give me a few tips to help me get started with the development of the fifo module? Thank you, Max
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120121/de2151ec/attachment.html>

From nginx-forum at nginx.us  Sat Jan 21 17:40:49 2012
From: nginx-forum at nginx.us (takigama)
Date: Sat, 21 Jan 2012 12:40:49 -0500
Subject: Reading from a named pipe / fifo
In-Reply-To: <E1Rodnb-00028Q-L0.nginxyz-mail-ru@f167.mail.ru>
References: <E1Rodnb-00028Q-L0.nginxyz-mail-ru@f167.mail.ru>
Message-ID: <a85afa7db22ef545c589aff24206d6d6.NginxMailingListEnglish@forum.nginx.org>

Max Wrote:
-------------------------------------------------------
> Unfortunately,
> the other app can provide the data only through a
> fifo and there's no easy way to change it due to
> licensing, so writing a fifo module for nginx
> would be the easier and cleaner solution. Igor and
> other developers, could you please give me a few
> tips to help me get started with the development
> of the fifo module? Thank you,
> Max

Personally, the way i'd do it is to write a bit of php (or python, perl,
etc) called via fast cgi to read your fifo (be mindful of the blocking
nature of pipes which can cause such things to hang waiting for input).

but,
http://squirrelshaterobots.com/programming/php/building-a-queue-server-in-php-part-3-accepting-input-from-named-pipes/
is an example of reading a fifo from php. And it would be quite easy to
do something like this (kinda off the top of my head):

        location /location/of/my/fifo.html {
                include /etc/nginx/fastcgi_params;
                fastcgi_pass
unix:/var/run/php-fastcgi/php-fastcgi.socket;
                fastcgi_param SCRIPT_FILENAME
/path/to/local/php/that/reads/pipe.php;
        }

within the nginx config...

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221499,221507#msg-221507


From nginx-forum at nginx.us  Sat Jan 21 17:45:02 2012
From: nginx-forum at nginx.us (moon35)
Date: Sat, 21 Jan 2012 12:45:02 -0500
Subject: HttpLimitZoneModule + Varnish (frontend)
Message-ID: <2aa280cdee23b7e15299a24be13bdfac.NginxMailingListEnglish@forum.nginx.org>

Hello,

I'm having some troubles with the Module LimitZone. 
http://wiki.nginx.org/HttpLimitZoneModule

I want to limit simultaneous connections per client (IP) with
"limit_conn" but it doesn't work with Varnish-cache which is listening
to port 80 and NginX: 8080 

To do some tests I put limit_conn to 1 and I tried to download two files
from my server
If I write this on my browser adress bar :
 -> http://mydomain.tld:8080/file_to_dl.ext
it works of course (I got the error 503) but this :
 -> http://mydomain.tld/file_to_dl.ext
doesn't work: I can download more than 1 file.

I'm using php-fpm and $_SERVER['remote_addr'] give the right Client IP

Does anyone have a solution to my problem ? 
Thanks in advance.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221508,221508#msg-221508


From lists at ruby-forum.com  Sat Jan 21 20:06:18 2012
From: lists at ruby-forum.com (Ian M.)
Date: Sat, 21 Jan 2012 21:06:18 +0100
Subject: Nginx causing high IO
Message-ID: <b4f489c5efec5de981ada30642fc5d1a@ruby-forum.com>

Hello ladies and gents,

I am using nginx to serve video on our video sharing platform. We are
seeing high IO from nginx which is causing extremely slow responsiveness
from the servers. I have tried changing the number of child nginx
services to four but still the services go into a D state.

Image: http://oi41.tinypic.com/214pq9.jpg
Image: http://oi43.tinypic.com/244slkx.jpg

My servers are running centos 5 32bit.

Here's how my configs look: http://tinypaste.com/bcc99261

There's some more to the configs but the rest is related to the script
that we're running to operate our website. Anyhow, any suggested would
be greatly appreciated as I'm not very familiar with nginx.

Thank you

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


From nginxyz at mail.ru  Sun Jan 22 01:54:08 2012
From: nginxyz at mail.ru (=?UTF-8?B?TWF4?=)
Date: Sun, 22 Jan 2012 05:54:08 +0400
Subject: Reading from a named pipe / fifo
In-Reply-To: <a85afa7db22ef545c589aff24206d6d6.NginxMailingListEnglish@forum.nginx.org>
References: <E1Rodnb-00028Q-L0.nginxyz-mail-ru@f167.mail.ru>
	<a85afa7db22ef545c589aff24206d6d6.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <E1RomdE-0006Hz-CH.nginxyz-mail-ru@f200.mail.ru>


21 ?????? 2012, 21:41 ?? "takigama" <nginx-forum at nginx.us>:
> 
> Personally, the way i'd do it is to write a bit of php (or python, perl,
> etc) called via fast cgi [...]

Thanks for the suggestion, but using FastCGI is not an option for
many reasons. The most elegant solution I could come up with
without hacking nginx involves embedded Perl like this:

===<FIFOReader.pm>===
use nginx;

sub handler {
    my $r = shift;
    alarm 5;
    open(FIFO, "</path/to/fifo") or return 444;
    while (<FIFO>) {
        $r->print($_);
    }
    $r->rflush();
    close(FIFO);
    return OK;
}
===</FIFOReader.pm>===

===<nginx.conf>===
http {
  perl_modules perl.modules/;
  perl_require FIFOReader.pm;
  
  server {
    location ~ ^/readfifo {
      ssi on;
      perl FIFOReader::handler;
    }
  }
}
===</nginx.conf>===

===<readfifo>===
<!-- # perl sub="FIFOReader::handler" -->
===</readfifo>===

FIFOReader.pm is located in $nginx_conf_dir/perl.modules/
readfifo is located in $document_root/

In the FIFOReader Perl module I first set the alarm to allow
the blocking open() call 5 seconds to open the FIFO. After
5 seconds the blocking open() call gets interrupted and
nginx returns code 444, which tears down the connection
without sending any kind of response code, so there is
no need to check for response codes - the client gets
either nothing or the data from the FIFO.

If the blocking open() call succeeds before the timeout,
the alarm is turned off and the data is read from
the FIFO and passed on to the client, after which the
FIFO is closed and the OK code is returned by nginx.

There is no additional alarm to prevent read timeouts
(before while (<FIFO>)) because that is left up to the
client by design.

Another side benefit of this approach is that it doesn't
require additional processes (such as FastCGI wrappers),
so there is no need to supervise additional processes to
make sure they're around. The embedded Perl module
will work for as long as nginx is running.

I have started working on a FIFO module to allow per-location
FIFO opening and reading. If I get it to production quality I'll
post my patches to the list.

Max

From nginxyz at mail.ru  Sun Jan 22 01:57:46 2012
From: nginxyz at mail.ru (=?UTF-8?B?TWF4?=)
Date: Sun, 22 Jan 2012 05:57:46 +0400
Subject: Reading from a named pipe / fifo (corrected)
In-Reply-To: <a85afa7db22ef545c589aff24206d6d6.NginxMailingListEnglish@forum.nginx.org>
References: <E1Rodnb-00028Q-L0.nginxyz-mail-ru@f167.mail.ru>
	<a85afa7db22ef545c589aff24206d6d6.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <E1Romgk-0003rw-6Z.nginxyz-mail-ru@f68.mail.ru>


21 ?????? 2012, 21:41 ?? "takigama" <nginx-forum at nginx.us>:
> 
> Personally, the way i'd do it is to write a bit of php (or python, perl,
> etc) called via fast cgi [...]

Thanks for the suggestion, but using FastCGI is not an option for
many reasons. The most elegant solution I could come up with
without hacking nginx involves embedded Perl like this:

===<FIFOReader.pm>===
package FIFOReader;

use nginx;

sub handler {
    my $r = shift;
    alarm 5;
    open(FIFO, "</path/to/fifo") or return 444;
    while (<FIFO>) {
        $r->print($_);
    }
    $r->rflush();
    close(FIFO);
    return OK;
}

1;
__END__
===</FIFOReader.pm>===

===<nginx.conf>===
http {
  perl_modules perl.modules/;
  perl_require FIFOReader.pm;
  
  server {
    location ~ ^/readfifo {
      ssi on;
      perl FIFOReader::handler;
    }
  }
}
===</nginx.conf>===

===<readfifo>===
<!-- # perl sub="FIFOReader::handler" -->
===</readfifo>===

FIFOReader.pm is located in $nginx_conf_dir/perl.modules/
readfifo is located in $document_root/

In the FIFOReader Perl module I first set the alarm to allow
the blocking open() call 5 seconds to open the FIFO. After
5 seconds the blocking open() call gets interrupted and
nginx returns code 444, which tears down the connection
without sending any kind of response code, so there is
no need to check for response codes - the client gets
either nothing or the data from the FIFO.

If the blocking open() call succeeds before the timeout,
the alarm is turned off and the data is read from
the FIFO and passed on to the client, after which the
FIFO is closed and the OK code is returned by nginx.

There is no additional alarm to prevent read timeouts
(before while (<FIFO>)) because that is left up to the
client by design.

Another side benefit of this approach is that it doesn't
require additional processes (such as FastCGI wrappers),
so there is no need to supervise additional processes to
make sure they're around. The embedded Perl module
will work for as long as nginx is running.

I have started working on a FIFO module to allow per-location
FIFO opening and reading. If I get it to production quality I'll
post my patches to the list.

Max

From nginxyz at mail.ru  Sun Jan 22 02:04:04 2012
From: nginxyz at mail.ru (=?UTF-8?B?TWF4?=)
Date: Sun, 22 Jan 2012 06:04:04 +0400
Subject: Reading from a named pipe / fifo (final version)
In-Reply-To: <a85afa7db22ef545c589aff24206d6d6.NginxMailingListEnglish@forum.nginx.org>
References: <E1Rodnb-00028Q-L0.nginxyz-mail-ru@f167.mail.ru>
	<a85afa7db22ef545c589aff24206d6d6.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <E1Rommq-0001dm-DL.nginxyz-mail-ru@f31.mail.ru>


21 ?????? 2012, 21:41 ?? "takigama" <nginx-forum at nginx.us>:
> 
> Personally, the way i'd do it is to write a bit of php (or python, perl,
> etc) called via fast cgi [...]

Thanks for the suggestion, but using FastCGI is not an option for
many reasons. The most elegant solution I could come up with
without hacking nginx involves embedded Perl like this:

===<FIFOReader.pm>===
package FIFOReader;

use nginx;

sub handler {
    my $r = shift;
    alarm 5;
    open(FIFO, "</path/to/fifo") or return 444;
    alarm 0;
    while (<FIFO>) {
        $r->print($_);
    }
    $r->rflush();
    close(FIFO);
    return OK;
}

1;
__END__
===</FIFOReader.pm>===

===<nginx.conf>===
http {
  perl_modules perl.modules/;
  perl_require FIFOReader.pm;
  
  server {
    location ~ ^/readfifo {
      ssi on;
      perl FIFOReader::handler;
    }
  }
}
===</nginx.conf>===

===<readfifo>===
<!-- # perl sub="FIFOReader::handler" -->
===</readfifo>===

FIFOReader.pm is located in $nginx_conf_dir/perl.modules/
readfifo is located in $document_root/

In the FIFOReader Perl module I first set the alarm to allow
the blocking open() call 5 seconds to open the FIFO. After
5 seconds the blocking open() call gets interrupted and
nginx returns code 444, which tears down the connection
without sending any kind of response code, so there is
no need to check for response codes - the client gets
either nothing or the data from the FIFO.

If the blocking open() call succeeds before the timeout,
the alarm is turned off and the data is read from
the FIFO and passed on to the client, after which the
FIFO is closed and the OK code is returned by nginx.

There is no additional alarm to prevent read timeouts
(before while (<FIFO>)) because that is left up to the
client by design.

Another side benefit of this approach is that it doesn't
require additional processes (such as FastCGI wrappers),
so there is no need to supervise additional processes to
make sure they're around. The embedded Perl module
will work for as long as nginx is running.

I have started working on a FIFO module to allow per-location
FIFO opening and reading. If I get it to production quality I'll
post my patches to the list.

Max

From nginx-forum at nginx.us  Sun Jan 22 02:20:35 2012
From: nginx-forum at nginx.us (DavidJS)
Date: Sat, 21 Jan 2012 21:20:35 -0500
Subject: Nginx causing high IO
In-Reply-To: <b4f489c5efec5de981ada30642fc5d1a@ruby-forum.com>
References: <b4f489c5efec5de981ada30642fc5d1a@ruby-forum.com>
Message-ID: <6d0cc015506aa63605a44dd7c67fd710.NginxMailingListEnglish@forum.nginx.org>

Which video sharing platform do you have? We are building a massive
network down in miami to serv a few million cable boxes at a time. glad
to help out if i can.
Whats the hardware looking like for the server?


Our Config Pulled from a beta node.
### User node24 in Beta Cluster 3
user node24;
worker_processes 54;
pid /var/run/nginx.pid;
  error_log   /logserver/nginx/cluster/node34/logs/error.log;
events
{

    worker_connections 65535;
     multi_accept on;
}

http
{
    ##
    # Changed. 
    ##
    sendfile on;
    tcp_nopush on;
    tcp_nodelay on;
    keepalive_timeout 9;
    types_hash_max_size 2048;
    # server_tokens off;
    # server_names_hash_bucket_size 64;
    # server_name_in_redirect off;

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221510,221515#msg-221515


From brian at akins.org  Sun Jan 22 02:26:05 2012
From: brian at akins.org (Brian Akins)
Date: Sat, 21 Jan 2012 21:26:05 -0500
Subject: Reading from a named pipe / fifo (final version)
In-Reply-To: <E1Rommq-0001dm-DL.nginxyz-mail-ru@f31.mail.ru>
References: <E1Rodnb-00028Q-L0.nginxyz-mail-ru@f167.mail.ru>
	<a85afa7db22ef545c589aff24206d6d6.NginxMailingListEnglish@forum.nginx.org>
	<E1Rommq-0001dm-DL.nginxyz-mail-ru@f31.mail.ru>
Message-ID: <D74678F8-ACD6-47A6-A04E-C408A816C8E9@akins.org>


On Jan 21, 2012, at 9:04 PM, Max wrote:
> 
> Thanks for the suggestion, but using FastCGI is not an option for
> many reasons. The most elegant solution I could come up with
> without hacking nginx involves embedded Perl like this:

Won't embedded perl block in this case? I'm not as familiar with it. If it does, this seems like a sub-optimal solution. FastCGI would not block on the nginx side. It's pretty trivial to manage fastcgi processes.

It doesn't have to be fastcgi, it could be a min-http server uwsgi or whatever or is there some restriction that nginx must talk directly to the fifo?



From nginxyz at mail.ru  Sun Jan 22 03:35:53 2012
From: nginxyz at mail.ru (=?UTF-8?B?TWF4?=)
Date: Sun, 22 Jan 2012 07:35:53 +0400
Subject: Reading from a named pipe / fifo (final version)
In-Reply-To: <D74678F8-ACD6-47A6-A04E-C408A816C8E9@akins.org>
References: <E1Rommq-0001dm-DL.nginxyz-mail-ru@f31.mail.ru>
	<E1Rodnb-00028Q-L0.nginxyz-mail-ru@f167.mail.ru>
	<D74678F8-ACD6-47A6-A04E-C408A816C8E9@akins.org>
Message-ID: <E1RooDh-0003bd-Ak.nginxyz-mail-ru@f222.mail.ru>


22 ?????? 2012, 06:26 ?? Brian Akins <brian at akins.org>:

> Won't embedded perl block in this case? I'm not as familiar with it. If it
> does, this seems like a sub-optimal solution. FastCGI would not block on the
> nginx side. It's pretty trivial to manage fastcgi processes.

It will block, but not indefinitely, only for as long as I want it to block.
That's why I used alarm to give open() 5 seconds to succeed - in case
it fails to open the FIFO within 5 seconds, it gets interrupted
(through SIGALRM) and since it fails the "or return 444" part
returns control to nginx to close the client's connection.

However, if the open() call succeeds within 5 seconds, the alarm
gets turned off (alarm 0) and then the FIFO is read from.

> It doesn't have to be fastcgi, it could be a min-http server uwsgi or whatever
> or is there some restriction that nginx must talk directly to the fifo?

Exactly, there are strict MAC (Mandatory Access Control) restrictions
in place. Even if there weren't, I'd still prefer the direct approach, it's
much simpler and cleaner, IMHO. Why would you want to use another
wrapper process and yet another process to supervise the wrapper
process and yet another process to read the FIFO and at least
another pair of sockets when you don't have to?

Max

From lists at ruby-forum.com  Sun Jan 22 18:53:56 2012
From: lists at ruby-forum.com (Ian M.)
Date: Sun, 22 Jan 2012 19:53:56 +0100
Subject: Nginx causing high IO
In-Reply-To: <6d0cc015506aa63605a44dd7c67fd710.NginxMailingListEnglish@forum.nginx.org>
References: <b4f489c5efec5de981ada30642fc5d1a@ruby-forum.com>
	<6d0cc015506aa63605a44dd7c67fd710.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <ead3457555a21592b09b319d0cbf92ae@ruby-forum.com>

DavidJS wrote in post #1041997:
> Which video sharing platform do you have? We are building a massive
> network down in miami to serv a few million cable boxes at a time. glad
> to help out if i can.
> Whats the hardware looking like for the server?

Hello David,

The hardware is the following.

E3-1270, 8GB Ram and 2x2TB HDD Raid 0 all over gbit connection.

We are using a custom built perl script for this. It was working nicely 
before on our old servers. Any ideas? I don't see anything obvious in 
your configs that could help me. I know your workers are pretty high but 
in my case I don't need that many. This is very frustrating because I'm 
seeing poor performance because of this.

Any help would be greatly appreciated :)

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


From Richard.Kearsley at m247.com  Sun Jan 22 19:39:54 2012
From: Richard.Kearsley at m247.com (Richard Kearsley)
Date: Sun, 22 Jan 2012 19:39:54 +0000
Subject: Nginx causing high IO
In-Reply-To: <ead3457555a21592b09b319d0cbf92ae@ruby-forum.com>
References: <b4f489c5efec5de981ada30642fc5d1a@ruby-forum.com>
	<6d0cc015506aa63605a44dd7c67fd710.NginxMailingListEnglish@forum.nginx.org>,
	<ead3457555a21592b09b319d0cbf92ae@ruby-forum.com>
Message-ID: <D3CC923C-9217-4F07-9A8C-35EA28970DFF@m247.com>

What disks did you have in old servers? The reason for high io is usually disks, but I have had it caused by cheap network cards too!

Also what does your perl script do exactly?

Sent from my iPhone

On 22 Jan 2012, at 18:54, "Ian M." <lists at ruby-forum.com> wrote:

> DavidJS wrote in post #1041997:
>> Which video sharing platform do you have? We are building a massive
>> network down in miami to serv a few million cable boxes at a time. glad
>> to help out if i can.
>> Whats the hardware looking like for the server?
> 
> Hello David,
> 
> The hardware is the following.
> 
> E3-1270, 8GB Ram and 2x2TB HDD Raid 0 all over gbit connection.
> 
> We are using a custom built perl script for this. It was working nicely 
> before on our old servers. Any ideas? I don't see anything obvious in 
> your configs that could help me. I know your workers are pretty high but 
> in my case I don't need that many. This is very frustrating because I'm 
> seeing poor performance because of this.
> 
> Any help would be greatly appreciated :)
> 
> -- 
> Posted via http://www.ruby-forum.com/.
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From brian at akins.org  Sun Jan 22 20:50:27 2012
From: brian at akins.org (Brian Akins)
Date: Sun, 22 Jan 2012 15:50:27 -0500
Subject: Reading from a named pipe / fifo (final version)
In-Reply-To: <E1RooDh-0003bd-Ak.nginxyz-mail-ru@f222.mail.ru>
References: <E1Rommq-0001dm-DL.nginxyz-mail-ru@f31.mail.ru>
	<E1Rodnb-00028Q-L0.nginxyz-mail-ru@f167.mail.ru>
	<D74678F8-ACD6-47A6-A04E-C408A816C8E9@akins.org>
	<E1RooDh-0003bd-Ak.nginxyz-mail-ru@f222.mail.ru>
Message-ID: <05BE587F-CEF7-4BC9-9D03-70C5C04924DE@akins.org>


On Jan 21, 2012, at 10:35 PM, Max wrote:

> 
> It will block, but not indefinitely, only for as long as I want it to block.
> That's why I used alarm to give open() 5 seconds to succeed
> 


5 seconds can be an eternity in internet time.

> 
> Exactly, there are strict MAC (Mandatory Access Control) restrictions
> in place.

You can use MAC with a fastcgi script.

> Even if there weren't, I'd still prefer the direct approach, it's
> much simpler and cleaner, IMHO.

I guess we'll just have to disagree about that. Doing things to make nginx block for seconds at a time on purpose seems much worse than managing a simple fastcgi script. IMNSHO.


--Brian

From nginx-forum at nginx.us  Sun Jan 22 21:06:10 2012
From: nginx-forum at nginx.us (DavidJS)
Date: Sun, 22 Jan 2012 16:06:10 -0500
Subject: Nginx causing high IO
In-Reply-To: <D3CC923C-9217-4F07-9A8C-35EA28970DFF@m247.com>
References: <D3CC923C-9217-4F07-9A8C-35EA28970DFF@m247.com>
Message-ID: <6999bd35206d4838798b71071c3cb904.NginxMailingListEnglish@forum.nginx.org>

Can you contact me outside the forum?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221510,221526#msg-221526


From mdounin at mdounin.ru  Sun Jan 22 21:43:40 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 23 Jan 2012 01:43:40 +0400
Subject: HttpLimitZoneModule + Varnish (frontend)
In-Reply-To: <2aa280cdee23b7e15299a24be13bdfac.NginxMailingListEnglish@forum.nginx.org>
References: <2aa280cdee23b7e15299a24be13bdfac.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120122214340.GG67687@mdounin.ru>

Hello!

On Sat, Jan 21, 2012 at 12:45:02PM -0500, moon35 wrote:

> Hello,
> 
> I'm having some troubles with the Module LimitZone. 
> http://wiki.nginx.org/HttpLimitZoneModule
> 
> I want to limit simultaneous connections per client (IP) with
> "limit_conn" but it doesn't work with Varnish-cache which is listening
> to port 80 and NginX: 8080 
> 
> To do some tests I put limit_conn to 1 and I tried to download two files
> from my server
> If I write this on my browser adress bar :
>  -> http://mydomain.tld:8080/file_to_dl.ext
> it works of course (I got the error 503) but this :

So nginx works as intended.

>  -> http://mydomain.tld/file_to_dl.ext
> doesn't work: I can download more than 1 file.
> 
> I'm using php-fpm and $_SERVER['remote_addr'] give the right Client IP
> 
> Does anyone have a solution to my problem ? 

Likely reason is that Varnish either doesn't pass your second 
request (or doesn't pass your requests at all, if it has the file 
in question cached) or it downloads the file faster that you are 
able to open second connection.  This behaviour is expected for a 
frontend, but it breaks limiting you are trying to do on a backend 
(nginx).

I doubt you'll be able solve this.  You need frontend to do the 
limiting, i.e. either use nginx as a frontend, or use similar 
functionality in Varnish (no idea if one is available).

Maxim Dounin


From nginx-forum at nginx.us  Mon Jan 23 14:41:57 2012
From: nginx-forum at nginx.us (moon35)
Date: Mon, 23 Jan 2012 09:41:57 -0500
Subject: HttpLimitZoneModule + Varnish (frontend)
In-Reply-To: <20120122214340.GG67687@mdounin.ru>
References: <20120122214340.GG67687@mdounin.ru>
Message-ID: <f93512b17c2ca36bfd392eaecbb5645a.NginxMailingListEnglish@forum.nginx.org>

Thanks a lot for your reply.
Unfortunately, according to what I read (on the internet) there is no
way to limit connections per client with varnish so I think I will
simply remove it.
Thanks again.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221508,221541#msg-221541


From nginx-forum at nginx.us  Mon Jan 23 23:00:20 2012
From: nginx-forum at nginx.us (gtuhl)
Date: Mon, 23 Jan 2012 18:00:20 -0500
Subject: Nginx as Load Balancer Connection Issues
In-Reply-To: <20451453746402827e0cb164c69a780b.NginxMailingListEnglish@forum.nginx.org>
References: <20451453746402827e0cb164c69a780b.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <b7d417f910acb59c6993e690cdd42a2f.NginxMailingListEnglish@forum.nginx.org>

gtuhl Wrote:
-------------------------------------------------------
> Initially we were seeing a ton of "connect()
> failed (110: Connection timed out)", 1 every
> couple seconds.  I added these to sysctl.conf and
> that seemed to solve the problem:
> 
> net.ipv4.tcp_syncookies = 1
> net.ipv4.tcp_fin_timeout = 20    
> net.ipv4.tcp_max_syn_backlog = 20480
> net.core.netdev_max_backlog = 4096
> net.ipv4.tcp_max_tw_buckets = 400000
> net.core.somaxconn = 4096
> 
> Now things generally run fine but every once in
> awhile we get a huge burst of "upstream
> prematurely closed connection while reading
> response header from upstream" followed by a "no
> live upstreams".  Again, no apparent load on the
> machines involved.  These bursts only last a
> minute or so.  We also still get an occasional
> "connect() failed (110: Connection timed out)" but
> they are far less frequent, perhaps 1 or 2 per
> hour.
> 

On looking at this again recently, we made two adjustments that
eliminated the connection issues completely:

net.nf_conntrack_max = 262144
net.ipv4.ip_local_port_range = 1024  65000

After making those two changes things became quite stable.  However, we
still have massive numbers of TIME_WAIT connections both on the nginx
machine and on the upstream apache machines.

The nginx machine is accepting roughly 1000 requests/s, and has 40,000
connections in TIME_WAIT.
The apache machines are each accepting roughly 250 requests/s, and have
15,000 connections in TIME_WAIT.

We tried setting net.ipv4.tcp_tw_reuse to 1 and restarting networking. 
That did not cause any trouble, but also didn't drop the TIME_WAIT
count.  I have read that net.ipv4.tcp_tw_recycle is dangerous but we may
try that if others have had good experiences.

Is there a way to have these cleaned up more quickly?  My concern is
that even with the expanded ip_local_port_range 40k is cutting it rather
close.  Before we bumped ip_local_port_range the whole system was
falling down right as the TIME_WAIT count approached 32k.  Is it normal
for nginx to cause this many TIME_WAIT connections?  If we're only doing
1k requests/s and nearly exhausting the available port range what would
sites with heavier volume do?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220894,221550#msg-221550


From nginx-forum at nginx.us  Tue Jan 24 00:19:36 2012
From: nginx-forum at nginx.us (Russell Mann)
Date: Mon, 23 Jan 2012 19:19:36 -0500
Subject: nginx serving large-ish static files on VM
Message-ID: <f9cbc71f748059f2c85462e2873c2c2a.NginxMailingListEnglish@forum.nginx.org>

I'm serving some 10-30mb files at peak times up to 250-300 requests per
second, if I'm reading these RDD'd /status logs correctly. These are on
VPS servers, which may mean some network settings aren't available, and
may mean disk IO is an issue.

It mostly works, but sometimes the massive simultaneous connections
overwhelms the system, and I'm wondering if anyone here is doing
something like that and has suggestions for better configs.

What's the proper use of sendfile in this instance?  I've read
conflicting information.  How about output_buffers?  tcp_delay,
tcp_nopush?

Thank you,

Russell

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221551,221551#msg-221551


From stef at scaleengine.com  Tue Jan 24 07:55:35 2012
From: stef at scaleengine.com (Stefan Caunter)
Date: Tue, 24 Jan 2012 02:55:35 -0500
Subject: HttpLimitZoneModule + Varnish (frontend)
In-Reply-To: <20120122214340.GG67687@mdounin.ru>
References: <2aa280cdee23b7e15299a24be13bdfac.NginxMailingListEnglish@forum.nginx.org>
	<20120122214340.GG67687@mdounin.ru>
Message-ID: <CALdpnryopk_UgPS0xKdxgZwy9A3BTCr_8pzjzq7FLi_2HTG-jg@mail.gmail.com>

On Sun, Jan 22, 2012 at 4:43 PM, Maxim Dounin <mdounin at mdounin.ru> wrote:
> Hello!
>
> On Sat, Jan 21, 2012 at 12:45:02PM -0500, moon35 wrote:
>
>> Hello,
>>
>> I'm having some troubles with the Module LimitZone.
>> http://wiki.nginx.org/HttpLimitZoneModule
>>
>> I want to limit simultaneous connections per client (IP) with
>> "limit_conn" but it doesn't work with Varnish-cache which is listening
>> to port 80 and NginX: 8080
>>
>> To do some tests I put limit_conn to 1 and I tried to download two files
>> from my server
>> If I write this on my browser adress bar :
>> ?-> http://mydomain.tld:8080/file_to_dl.ext
>> it works of course (I got the error 503) but this :
>
> So nginx works as intended.
>
>> ?-> http://mydomain.tld/file_to_dl.ext
>> doesn't work: I can download more than 1 file.
>>
>> I'm using php-fpm and $_SERVER['remote_addr'] give the right Client IP
>>
>> Does anyone have a solution to my problem ?
>
> Likely reason is that Varnish either doesn't pass your second
> request (or doesn't pass your requests at all, if it has the file
> in question cached) or it downloads the file faster that you are
> able to open second connection. ?This behaviour is expected for a
> frontend, but it breaks limiting you are trying to do on a backend
> (nginx).
>
> I doubt you'll be able solve this. ?You need frontend to do the
> limiting, i.e. either use nginx as a frontend, or use similar
> functionality in Varnish (no idea if one is available).

For this kind of perimeter rate limiting technique, you would
increment a header value in varnish.
Set a custom header in vcl_recv, qualified or limited by the specific
directory or valuable asset, for the external IP to value 1 for the
first request, which is allowed, and add a test to see if that flag
has been set; add vcl_error logic to redirect with message, or send
403, if there are subsequent requests, i.e. if the custom header > 1,
or 100, or whatever you set as a limit threshold.

Stefan Caunter
http://scaleengine.com/contact

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


From kurienmatheww at gmail.com  Tue Jan 24 11:10:35 2012
From: kurienmatheww at gmail.com (Kurien Mathew)
Date: Tue, 24 Jan 2012 12:10:35 +0100
Subject: nginx Window native
In-Reply-To: <11a5f1cdd83bd25c9ce006182286521f.NginxMailingListEnglish@forum.nginx.org>
References: <CAMwgYSuMbNDpPROe8XR2oH4KLK-WpLdpwJmfa2=sdiiCU3Lt6A@mail.gmail.com>
	<11a5f1cdd83bd25c9ce006182286521f.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CAMwgYSvJLc2PFj6QZDdrtCGsfYdNNOy16FBr06uKMfoOgNiJAQ@mail.gmail.com>

Hello,

I tested the Cywin version along with the following additional
configuration settings:

worker_processes  5;
worker_rlimit_nofile 3072;

events {
    worker_connections  1024;
    accept_mutex off;
    multi_accept on;
}

The exact effect of the above settings on the functioning of nginx is not
well understood. But with this configuration I find that multiple workers
do the work and this improves scalability. One problem that I discovered
was that sometimes a connection is not accepted and the following error is
emitted:

2012/01/24 11:45:16 [alert] 15964#0: accept() failed (14: Bad address)

For eg with 2000 concurrent connections about 15 fail to be accepted. When
the client shuts down after waiting, the server sockets go into CLOSE_WAIT
state. They seem to remain that way forever.

Is anyone aware of a way to overcome this problem?

Thanks
Kurien


On Fri, Dec 9, 2011 at 10:38 AM, Ensiferous <nginx-forum at nginx.us> wrote:

> Additionally the Windows version of the event polling which enables high
> performance on unix is called IOCP and is not supported last I checked.
> So you would basically be using select().
>
> Windows support is very preliminary and while I can't speak for the
> development team, I don't think it's supposed to be used for anything
> other than development or testing.
>
> -Martin Fjordvald
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,219822,219825#msg-219825
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120124/4a76e163/attachment.html>

From nginx-forum at nginx.us  Tue Jan 24 11:37:12 2012
From: nginx-forum at nginx.us (mangler)
Date: Tue, 24 Jan 2012 06:37:12 -0500
Subject: proxy_pass to upstream with client certificate
Message-ID: <8322755aaaaa737176a515a38ec3feec.NginxMailingListEnglish@forum.nginx.org>

Is it possible to configure nginx to send a client certificate to
upstream?

I'd like the flow to be:
HTTP to nginx which proxies to HTTPS://xyz/ that uses a client
certificate.

I don't want the end client to use any certificate what so ever
(preferably).

Regards,
Magnus J?derberg

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221563,221563#msg-221563


From mdounin at mdounin.ru  Tue Jan 24 11:46:41 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 24 Jan 2012 15:46:41 +0400
Subject: proxy_pass to upstream with client certificate
In-Reply-To: <8322755aaaaa737176a515a38ec3feec.NginxMailingListEnglish@forum.nginx.org>
References: <8322755aaaaa737176a515a38ec3feec.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120124114641.GF67687@mdounin.ru>

Hello!

On Tue, Jan 24, 2012 at 06:37:12AM -0500, mangler wrote:

> Is it possible to configure nginx to send a client certificate to
> upstream?
> 
> I'd like the flow to be:
> HTTP to nginx which proxies to HTTPS://xyz/ that uses a client
> certificate.
> 
> I don't want the end client to use any certificate what so ever
> (preferably).

No, it's not currently supported.

Maxim Dounin


From nginx-forum at nginx.us  Tue Jan 24 15:26:29 2012
From: nginx-forum at nginx.us (lockev3.0)
Date: Tue, 24 Jan 2012 10:26:29 -0500
Subject: Default_server catch all block not working
Message-ID: <4b142fa24c8249c90f455f6736fe5ccf.NginxMailingListEnglish@forum.nginx.org>

Hi, getting quite desperate at the moment about this.

I have a server serving several subdomains; actually proxying'em to
apache2. So I have main nginx vhost file with all server_name's added.
Such server names match the defined ServerName's and ServerAlias'es of
Apache.
Beside, I have one more file to  define a default_server catch all block
like follows

server {
        listen       80 default_server;
        server_name  _;
        error_log       /var/log/nginx/000default-error.log error;
        root /var/www/forvo.com/_templates;
        error_page 404 /404.html;
        location / {
                #return       444;
                return 404;
        }
        location = /404.html {
                internal;
        }
        #rewrite .* http://myfaultysubdomain.com permanent;
}


As soon as I enable this file, everything works as expected, however, in
less than one minute (my site is always being visited), all requests
(good subdomains and bad subdomains as well) start to get 503 response.
No log at all seems to change when this happens.

Any clue to what may be happening pleeeeeease ......it's been hours,
during days, trying to figure out what the h@$!! is going on

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221572,221572#msg-221572


From appa at perusio.net  Tue Jan 24 15:33:55 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Tue, 24 Jan 2012 15:33:55 +0000
Subject: Default_server catch all block not working
In-Reply-To: <4b142fa24c8249c90f455f6736fe5ccf.NginxMailingListEnglish@forum.nginx.org>
References: <4b142fa24c8249c90f455f6736fe5ccf.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87pqe915ek.wl%appa@perusio.net>

On 24 Jan 2012 15h26 WET, nginx-forum at nginx.us wrote:

> Hi, getting quite desperate at the moment about this.
>
> I have a server serving several subdomains; actually proxying'em to
> apache2. So I have main nginx vhost file with all server_name's
> added.  Such server names match the defined ServerName's and
> ServerAlias'es of Apache.  Beside, I have one more file to define a
> default_server catch all block like follows
>
> server {
> listen       80 default_server;
> server_name  _;
> error_log       /var/log/nginx/000default-error.log error;
> root /var/www/forvo.com/_templates;
> error_page 404 /404.html;
> location / {
> #return       444;
> return 404;
> }
> location = /404.html {
> internal;
> }
> #rewrite .* http://myfaultysubdomain.com permanent;
> }

AFAICT there's nothing here that should cause the 503s. Do you have
any limits set, be it connection or rate limit?

--- appa





>
> As soon as I enable this file, everything works as expected,
> however, in less than one minute (my site is always being visited),
> all requests (good subdomains and bad subdomains as well) start to
> get 503 response.  No log at all seems to change when this happens.
>
> Any clue to what may be happening pleeeeeease ......it's been hours,
> during days, trying to figure out what the h@$!! is going on
>
> Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,221572,221572#msg-221572
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From mit at stagename.com  Tue Jan 24 16:49:59 2012
From: mit at stagename.com (Mit Rowe)
Date: Tue, 24 Jan 2012 11:49:59 -0500
Subject: Default_server catch all block not working
In-Reply-To: <87pqe915ek.wl%appa@perusio.net>
References: <4b142fa24c8249c90f455f6736fe5ccf.NginxMailingListEnglish@forum.nginx.org>
	<87pqe915ek.wl%appa@perusio.net>
Message-ID: <CAFWPQt4FPPDZtFxtCjMOSVCWwss9CTx6iR0E0XVfMQCsxcQE_A@mail.gmail.com>

we might need to see your whole configuration to get the full picture, but
here's an initial thought...

if your apache does virtual hosting based on the Host: header (which seems
likely from your description), you may need to pass the Host: header along,
which you can do like this:

proxy_set_header Host $http_host;



On Tue, Jan 24, 2012 at 10:33 AM, Ant?nio P. P. Almeida <appa at perusio.net>wrote:

> On 24 Jan 2012 15h26 WET, nginx-forum at nginx.us wrote:
>
> > Hi, getting quite desperate at the moment about this.
> >
> > I have a server serving several subdomains; actually proxying'em to
> > apache2. So I have main nginx vhost file with all server_name's
> > added.  Such server names match the defined ServerName's and
> > ServerAlias'es of Apache.  Beside, I have one more file to define a
> > default_server catch all block like follows
> >
> > server {
> > listen       80 default_server;
> > server_name  _;
> > error_log       /var/log/nginx/000default-error.log error;
> > root /var/www/forvo.com/_templates;
> > error_page 404 /404.html;
> > location / {
> > #return       444;
> > return 404;
> > }
> > location = /404.html {
> > internal;
> > }
> > #rewrite .* http://myfaultysubdomain.com permanent;
> > }
>
> AFAICT there's nothing here that should cause the 503s. Do you have
> any limits set, be it connection or rate limit?
>
> --- appa
>
>
>
>
>
> >
> > As soon as I enable this file, everything works as expected,
> > however, in less than one minute (my site is always being visited),
> > all requests (good subdomains and bad subdomains as well) start to
> > get 503 response.  No log at all seems to change when this happens.
> >
> > Any clue to what may be happening pleeeeeease ......it's been hours,
> > during days, trying to figure out what the h@$!! is going on
> >
> > Posted at Nginx Forum:
> > http://forum.nginx.org/read.php?2,221572,221572#msg-221572
> >
> > _______________________________________________
> > nginx mailing list
> > nginx at nginx.org
> > http://mailman.nginx.org/mailman/listinfo/nginx
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>



-- 
Will 'Mit' Rowe
Stagename*
*1-866-326-3098
mit at stagename.com <josh at stagename.com>
www.stagename.com
Twitter: @stagename

*The information transmitted is intended only for the person or entity to
which it is addressed and may contain confidential and/or privileged
material. Any review, retransmission, dissemination or other use of this
information by persons or entities other than the intended recipient is
prohibited. If you received this transmission in error, please contact the
sender and delete all material contained herein from your computer.*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120124/a8fdc706/attachment.html>

From martinloy.uy at gmail.com  Tue Jan 24 17:33:28 2012
From: martinloy.uy at gmail.com (Martin Loy)
Date: Tue, 24 Jan 2012 15:33:28 -0200
Subject: proxy_pass to upstream with client certificate
In-Reply-To: <20120124114641.GF67687@mdounin.ru>
References: <8322755aaaaa737176a515a38ec3feec.NginxMailingListEnglish@forum.nginx.org>
	<20120124114641.GF67687@mdounin.ru>
Message-ID: <CAHbGeW5+W+DAmcHquj4eeiyL=ikKbhXVo+6H1Cqky-66oidnKg@mail.gmail.com>

alternative if you really need it, you can tunnel through ssh/vpn with
all the overhead involved

On Tue, Jan 24, 2012 at 9:46 AM, Maxim Dounin <mdounin at mdounin.ru> wrote:
> Hello!
>
> On Tue, Jan 24, 2012 at 06:37:12AM -0500, mangler wrote:
>
>> Is it possible to configure nginx to send a client certificate to
>> upstream?
>>
>> I'd like the flow to be:
>> HTTP to nginx which proxies to HTTPS://xyz/ that uses a client
>> certificate.
>>
>> I don't want the end client to use any certificate what so ever
>> (preferably).
>
> No, it's not currently supported.
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx



-- 
Nunca hubo un amigo que hiciese un favor a un enano, ni un enemigo que
le hiciese un mal, que no se viese recompensado por entero.


From nginx-forum at nginx.us  Tue Jan 24 17:57:48 2012
From: nginx-forum at nginx.us (lockev3.0)
Date: Tue, 24 Jan 2012 12:57:48 -0500
Subject: Default_server catch all block not working
In-Reply-To: <4b142fa24c8249c90f455f6736fe5ccf.NginxMailingListEnglish@forum.nginx.org>
References: <4b142fa24c8249c90f455f6736fe5ccf.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <669ff20b8722106d70658bca002eed4e.NginxMailingListEnglish@forum.nginx.org>

glubs !!! header host looks like "the missing culprit", I'll check it
out to confirm your nice guess Mit Rowe

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221572,221582#msg-221582


From nginx-forum at nginx.us  Tue Jan 24 17:59:54 2012
From: nginx-forum at nginx.us (ggrensteiner)
Date: Tue, 24 Jan 2012 12:59:54 -0500
Subject: Nginx as Load Balancer Connection Issues
In-Reply-To: <b7d417f910acb59c6993e690cdd42a2f.NginxMailingListEnglish@forum.nginx.org>
References: <20451453746402827e0cb164c69a780b.NginxMailingListEnglish@forum.nginx.org>
	<b7d417f910acb59c6993e690cdd42a2f.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <e32e3812bd6d67fa72e1bb9eefb4892f.NginxMailingListEnglish@forum.nginx.org>

net.ipv4.tcp_tw_recycle = 1

is what your looking for

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220894,221583#msg-221583


From nginx-forum at nginx.us  Tue Jan 24 18:05:24 2012
From: nginx-forum at nginx.us (locojohn)
Date: Tue, 24 Jan 2012 13:05:24 -0500
Subject: [ANNOUNCE] Tengine-1.2.2
In-Reply-To: <87aa5syd32.wl%appa@perusio.net>
References: <87aa5syd32.wl%appa@perusio.net>
Message-ID: <711f3a182e8e060292f254c22cc41f5a.NginxMailingListEnglish@forum.nginx.org>

Hello,

Joshua Zhu, I find http-concat module very useful indeed.   Except there
are a few minor issues that hopefully will be fixed in the new
version(s):

1. It is not possible to define files from alternate folder followed by
files defined by another folder, for example:

/jquery/??jquery.asmselect.css,ui.imageuploader.css,jquery.linkselect.css
   => works

but

/jquery/??jquery.asmselect.css,ui.imageuploader.css,jquery.linkselect.css,/site-default.css
 => doesn't work

(site-default.css resides in the root)

Suggestion is therefore if "/" is used as a prefix for consequent files,
then treat this component as absolute URL

2. Some CSS files may contain relative links to images.  However, when
these CSS files are concatenated using http-concat from a page residing
in a different folder, links to images will not of course work anymore,
as they become relative to the current page.   For example: 
"jquery-ui-1.8.17.custom.css" of the standard jquery library contains 8
links to images/* and of course they don't work.

We wrote our own PHP script to automatically parse CSS files and fix the
url's by using the following perl regex:

$path = dirname($e['uri']);
// example: url(relative/path/to/file) becomes:
url(/absolute/and/not/relative/path/to/file) within a CSS file
$buf =
preg_replace('/(:?\s*url\s*\()[\'"\s]*([^\/\'"].*)[\'"\s]*\)/isU', '$1'
. ($path == '/' ? '/' : $path . '/') . '$2)', $buf);

It doesn't seem to create a big processing overhead, since the files are
read from disk anyway, and this only does a replace operation in the
memory buffer.   what do you say?  would it be possible to implement?

Also, Antonio, you did a great job to distribute http-concat on git, but
the link to download the archive does not appear to have a .gz/.zip
extension, and, for example, in Gentoo distribution it creates problems
when using "unpack()" function in its ebuilds, as it doesn't know what
type of archive it is:

https://nodeload.github.com/perusio/nginx-http-concat/tarball/master

(doesn't end with ".gz")

Also, there is no version information.

Best regards

Andrejs

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221087,221584#msg-221584


From daniel.carrillo at gmail.com  Tue Jan 24 18:12:07 2012
From: daniel.carrillo at gmail.com (Daniel Carrillo)
Date: Tue, 24 Jan 2012 19:12:07 +0100
Subject: Possible memory leak nginx 1.1.12 with pcre and jit enabled
Message-ID: <CAK6GuySUx9KmvJ4Ugv9cmVcrHH5Uc2Z=oAPkpQ9Uj+UW9BTjnA@mail.gmail.com>

Hi all.

I compiled nginx against pcre 8.21 and jit enabled. With this scenario
in a few days the nginx processes are using more than 100MiB of RAM
each. The same scenario with nginx 1.1.12 compiled against the system
pcre library and jit disable (not available) the nginx processes are
not consuming more than 13MiB each (as expected).

CentOS 6.x
nginx 1.1.12
./configure --with-pcre=../pcre-8.21/ --with-pcre-jit
--with-http_stub_status_module --with-http_ssl_module
--without-select_module --without-http_ssi_module
--without-http_userid_module --without-poll_module
--without-http_ssi_module --without-http_autoindex_module
--without-http_geo_module --without-http_map_module
--without-http_split_clients_module --without-http_referer_module
--without-http_uwsgi_module --without-http_scgi_module
--without-http_memcached_module --without-http_limit_conn_module
--without-http_limit_req_module --without-http_browser_module
--without-http_upstream_ip_hash_module --without-http-cache
--without-mail_pop3_module --without-mail_imap_module
--without-mail_smtp_module

I'm not sure if the memory leak is related to nginx, pcre,
configuration or a mix. Please let me know if more information is
needed, configuration files, etc.

Thanks in advanced.
Kind Regards.


From andrey at xdel.ru  Tue Jan 24 18:12:56 2012
From: andrey at xdel.ru (Andrey Korolyov)
Date: Tue, 24 Jan 2012 21:12:56 +0300
Subject: Nginx as Load Balancer Connection Issues
In-Reply-To: <e32e3812bd6d67fa72e1bb9eefb4892f.NginxMailingListEnglish@forum.nginx.org>
References: <20451453746402827e0cb164c69a780b.NginxMailingListEnglish@forum.nginx.org>
	<b7d417f910acb59c6993e690cdd42a2f.NginxMailingListEnglish@forum.nginx.org>
	<e32e3812bd6d67fa72e1bb9eefb4892f.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CABYiri-Gwpm2m7pdgLSp0rWa1KxAipTTDH9Ca_xvU3_cszq-6Q@mail.gmail.com>

On Tue, Jan 24, 2012 at 9:59 PM, ggrensteiner <nginx-forum at nginx.us> wrote:
> net.ipv4.tcp_tw_recycle = 1
>
> is what your looking for
>
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220894,221583#msg-221583
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx

This may cause trouble if multiple clients trying to reach the server
over same NAT, so be careful. I have a negative experience even on ~
10 http reqs/min from NAT machine.


From nginx-forum at nginx.us  Tue Jan 24 18:23:38 2012
From: nginx-forum at nginx.us (gtuhl)
Date: Tue, 24 Jan 2012 13:23:38 -0500
Subject: Nginx as Load Balancer Connection Issues
In-Reply-To: <CABYiri-Gwpm2m7pdgLSp0rWa1KxAipTTDH9Ca_xvU3_cszq-6Q@mail.gmail.com>
References: <CABYiri-Gwpm2m7pdgLSp0rWa1KxAipTTDH9Ca_xvU3_cszq-6Q@mail.gmail.com>
Message-ID: <a812381f5ffe56a06d2b61547b2e1b0c.NginxMailingListEnglish@forum.nginx.org>

Andrey Korolyov Wrote:
-------------------------------------------------------
> On Tue, Jan 24, 2012 at 9:59 PM, ggrensteiner
> <nginx-forum at nginx.us> wrote:
> > net.ipv4.tcp_tw_recycle = 1
> >
> > is what your looking for
> >
> > Posted at Nginx Forum:
> http://forum.nginx.org/read.php?2,220894,221583#ms
> g-221583
> >
> > _______________________________________________
> > nginx mailing list
> > nginx at nginx.org
> > http://mailman.nginx.org/mailman/listinfo/nginx
> 
> This may cause trouble if multiple clients trying
> to reach the server
> over same NAT, so be careful. I have a negative
> experience even on ~
> 10 http reqs/min from NAT machine.
> 

This is what I had read everywhere as well, so I've been hesitant to try
it.  We definitely have a lot of users that would be coming at our
servers from the same buliding/NAT.  

Has anyone tried using "net.ipv4.tcp_tw_reuse = 1" in a larger
connection count environment before?

I have it enabled now, but it did not seem to have any impact on the
number of TIME_WAIT connections.  Does it wait until it actually needs
to reuse one (due to port exhaustion) before doing so?  Or should it be
keeping the number lower?

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220894,221587#msg-221587


From ne at vbart.ru  Tue Jan 24 19:17:40 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Tue, 24 Jan 2012 23:17:40 +0400
Subject: Possible memory leak nginx 1.1.12 with pcre and jit enabled
In-Reply-To: <CAK6GuySUx9KmvJ4Ugv9cmVcrHH5Uc2Z=oAPkpQ9Uj+UW9BTjnA@mail.gmail.com>
References: <CAK6GuySUx9KmvJ4Ugv9cmVcrHH5Uc2Z=oAPkpQ9Uj+UW9BTjnA@mail.gmail.com>
Message-ID: <201201242317.40871.ne@vbart.ru>

On Tuesday 24 January 2012 22:12:07 Daniel Carrillo wrote:
> Hi all.
> 
> I compiled nginx against pcre 8.21 and jit enabled. With this scenario
> in a few days the nginx processes are using more than 100MiB of RAM
> each. The same scenario with nginx 1.1.12 compiled against the system
> pcre library and jit disable (not available) the nginx processes are
> not consuming more than 13MiB each (as expected).
> 
> CentOS 6.x
> nginx 1.1.12
> ./configure --with-pcre=../pcre-8.21/ --with-pcre-jit
> --with-http_stub_status_module --with-http_ssl_module
> --without-select_module --without-http_ssi_module
> --without-http_userid_module --without-poll_module
> --without-http_ssi_module --without-http_autoindex_module
> --without-http_geo_module --without-http_map_module
> --without-http_split_clients_module --without-http_referer_module
> --without-http_uwsgi_module --without-http_scgi_module
> --without-http_memcached_module --without-http_limit_conn_module
> --without-http_limit_req_module --without-http_browser_module
> --without-http_upstream_ip_hash_module --without-http-cache
> --without-mail_pop3_module --without-mail_imap_module
> --without-mail_smtp_module
> 
> I'm not sure if the memory leak is related to nginx, pcre,
> configuration or a mix. Please let me know if more information is
> needed, configuration files, etc.
> 

Yes, please, provide your nginx config.

Which version is your system PCRE library? Does it leak when the "pcre_jit" 
directive is switched off?

 wbr, Valentin V. Bartenev


From appa at perusio.net  Tue Jan 24 19:51:25 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Tue, 24 Jan 2012 19:51:25 +0000
Subject: Default_server catch all block not working
In-Reply-To: <669ff20b8722106d70658bca002eed4e.NginxMailingListEnglish@forum.nginx.org>
References: <4b142fa24c8249c90f455f6736fe5ccf.NginxMailingListEnglish@forum.nginx.org>
	<669ff20b8722106d70658bca002eed4e.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87vco0vpz6.wl%appa@perusio.net>

On 24 Jan 2012 17h57 WET, nginx-forum at nginx.us wrote:

> glubs !!! header host looks like "the missing culprit", I'll check
> it out to confirm your nice guess Mit Rowe

444 closes the connection. Doesn't return any 503 code.

--- appa


From nginx-forum at nginx.us  Tue Jan 24 20:42:49 2012
From: nginx-forum at nginx.us (runesoerensen)
Date: Tue, 24 Jan 2012 15:42:49 -0500
Subject: Worker processes not shutting down
Message-ID: <1845550347547ea591de17585b45e02a.NginxMailingListEnglish@forum.nginx.org>

Following an upgrade from nginx 1.0.2 to 1.0.11 I've experienced a
problem with worker processes that are not shutting down. When looking
in the list of running processes the worker processes appears to be
shutting down but never actually exits.

I've been using the ubuntu nginx ppa's for maverick.

As the worker processes doesn't shut down nginx ends up consuming to
much memory and is ultimately unable to spawn new worker processes.

Below is a list of processes showing the problem:

www-data   560  0.5  4.2 376376 321760 ?       S    20:04   0:07 nginx:
worker process is shutting down
www-data   562  0.1  4.2 376376 321720 ?       S    20:04   0:01 nginx:
worker process is shutting down
www-data  1029  0.2  4.2 377712 323092 ?       S    20:10   0:02 nginx:
worker process is shutting down
www-data  1030  0.0  4.2 377712 322992 ?       S    20:10   0:00 nginx:
worker process is shutting down
www-data  1093  0.2  4.2 377972 323300 ?       S    20:11   0:02 nginx:
worker process is shutting down
www-data  1094  0.0  4.2 377972 323260 ?       S    20:11   0:00 nginx:
worker process is shutting down
www-data  1095  0.0  4.2 377972 323300 ?       S    20:11   0:00 nginx:
worker process is shutting down
www-data  1158  0.0  4.2 377972 323272 ?       S    20:12   0:00 nginx:
worker process is shutting down
www-data  1342  0.4  4.2 377972 323300 ?       S    20:14   0:03 nginx:
worker process is shutting down
www-data  1879  0.1  4.2 376928 322220 ?       S    20:21   0:00 nginx:
worker process is shutting down
www-data  1881  0.4  4.2 376928 322260 ?       S    20:21   0:01 nginx:
worker process is shutting down
www-data  2190  0.1  4.2 377972 323052 ?       S    20:25   0:00 nginx:
worker process is shutting down
www-data  2193  2.4  4.2 377972 323348 ?       S    20:25   0:04 nginx:
worker process is shutting down
www-data  2194  0.5  4.2 377972 323272 ?       S    20:25   0:01 nginx:
worker process is shutting down
www-data  2376  0.3  4.1 375884 321060 ?       S    20:28   0:00 nginx:
worker process
www-data  2377  0.5  4.1 375884 320320 ?       S    20:28   0:00 nginx:
worker process
www-data  2378  0.2  4.1 375884 319780 ?       S    20:28   0:00 nginx:
worker process
www-data  2379  4.0  4.1 375884 321192 ?       S    20:28   0:02 nginx:
worker process
ubuntu    2576  0.0  0.0   7964   900 pts/1    R+   20:29   0:00 grep
--color=auto nginx
root      3185  8.2  4.1 373960 318980 ?       Rs   Jan09 1854:37 nginx:
master process nginx
www-data 25687  0.1  4.2 377420 322804 ?       S    18:28   0:12 nginx:
worker process is shutting down
www-data 25688  0.1  4.2 377420 322768 ?       S    18:28   0:10 nginx:
worker process is shutting down
www-data 25689  0.2  4.2 377420 322804 ?       S    18:28   0:17 nginx:
worker process is shutting down
www-data 26682  0.1  4.2 377676 323076 ?       S    18:41   0:10 nginx:
worker process is shutting down
www-data 27287  0.2  4.2 377004 322344 ?       S    18:49   0:15 nginx:
worker process is shutting down
www-data 27289  0.0  4.2 377004 322276 ?       S    18:49   0:04 nginx:
worker process is shutting down
www-data 27290  0.1  4.2 377004 322356 ?       S    18:49   0:10 nginx:
worker process is shutting down
www-data 29293  0.0  4.2 376608 322116 ?       S    19:15   0:04 nginx:
worker process is shutting down
www-data 29619  0.0  4.2 376608 322076 ?       S    19:20   0:02 nginx:
worker process is shutting down
www-data 29620  0.1  4.2 376608 322116 ?       S    19:20   0:04 nginx:
worker process is shutting down
www-data 30116  0.0  4.2 376608 322068 ?       S    19:26   0:02 nginx:
worker process is shutting down
www-data 30497  0.0  4.2 376608 322064 ?       S    19:31   0:03 nginx:
worker process is shutting down
www-data 31235  0.1  4.2 376536 321904 ?       S    19:41   0:05 nginx:
worker process is shutting down
www-data 31817  0.0  4.2 376544 321896 ?       S    19:49   0:01 nginx:
worker process is shutting down
www-data 31938  0.1  4.2 376460 321848 ?       S    19:50   0:03 nginx:
worker process is shutting down
www-data 31939  0.0  4.2 376460 321724 ?       S    19:50   0:01 nginx:
worker process is shutting down
www-data 32368  0.5  4.2 376412 321800 ?       S    19:56   0:10 nginx:
worker process is shutting down

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221591,221591#msg-221591


From nginx-forum at nginx.us  Tue Jan 24 21:17:34 2012
From: nginx-forum at nginx.us (runesoerensen)
Date: Tue, 24 Jan 2012 16:17:34 -0500
Subject: Worker processes not shutting down
In-Reply-To: <1845550347547ea591de17585b45e02a.NginxMailingListEnglish@forum.nginx.org>
References: <1845550347547ea591de17585b45e02a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <ef983449aae19bd6001a7e209ca0ae78.NginxMailingListEnglish@forum.nginx.org>

The problem occurs when reloading the configuration - where nginx
usually exits the existing worker processes gracefully.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221591,221593#msg-221593


From nginx-forum at nginx.us  Wed Jan 25 08:49:40 2012
From: nginx-forum at nginx.us (evilsocket)
Date: Wed, 25 Jan 2012 03:49:40 -0500
Subject: Module API and srv structure initialization.
Message-ID: <3a83eac6e464a4bd300408e4a645f7b9.NginxMailingListEnglish@forum.nginx.org>

Hi guys, i'm new to this forum (already introduced myself ^^).

I'm writing a nginx handler module and i need to load a file in a server
directive inside the config, let's assume it's something like:

server {
 ...
 mymodule-ruleset /etc/nginx/mymodule-rules.json;
 ...
}

So i have a server structure like:

typedef struct {
 ...
 ngx_str_t ruleset;
 ...
}
ngx_mymodule_srv_conf_t;

I initialize the ruleset attribute inside the merge_srv_conf handler:

ngx_conf_merge_str_value( conf->ruleset_path, prev->ruleset_path,
DEFAULT_RULESET_PATH );

( Where DEFAULT_RULESET_PATH is a constant #defined my me ).

The problem is, i have to load this file only once for server directive,
but the user could ignore the mymodule-ruleset using its default value,
so i can't use an handler to load it (it could be never called).

Where's the best place to load this file? In the postconfiguration
handler ? If so, how do i retrieve the srv_conf_t inside the
postconfiguration handler ?

Thanks for your time, i'm new to nginx developing and i'm finding some
trouble to understand its structure.

PS: I've already read the guides specified in the wiki, but there's
nothing so specific.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221610,221610#msg-221610


From mdounin at mdounin.ru  Wed Jan 25 13:47:09 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Wed, 25 Jan 2012 17:47:09 +0400
Subject: Worker processes not shutting down
In-Reply-To: <1845550347547ea591de17585b45e02a.NginxMailingListEnglish@forum.nginx.org>
References: <1845550347547ea591de17585b45e02a.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120125134709.GQ67687@mdounin.ru>

Hello!

On Tue, Jan 24, 2012 at 03:42:49PM -0500, runesoerensen wrote:

> Following an upgrade from nginx 1.0.2 to 1.0.11 I've experienced a
> problem with worker processes that are not shutting down. When looking
> in the list of running processes the worker processes appears to be
> shutting down but never actually exits.
> 
> I've been using the ubuntu nginx ppa's for maverick.
> 
> As the worker processes doesn't shut down nginx ends up consuming to
> much memory and is ultimately unable to spawn new worker processes.
> 
> Below is a list of processes showing the problem:
> 
> www-data   560  0.5  4.2 376376 321760 ?       S    20:04   0:07 nginx:
> worker process is shutting down
> www-data   562  0.1  4.2 376376 321720 ?       S    20:04   0:01 nginx:
> worker process is shutting down

[...]

What does "nginx -V" show?  If there are any 3rd party 
modules/patches - do you see the same behaviour with vanilla 
nginx?

Maxim Dounin


From mdounin at mdounin.ru  Wed Jan 25 14:35:06 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Wed, 25 Jan 2012 18:35:06 +0400
Subject: Module API and srv structure initialization.
In-Reply-To: <3a83eac6e464a4bd300408e4a645f7b9.NginxMailingListEnglish@forum.nginx.org>
References: <3a83eac6e464a4bd300408e4a645f7b9.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120125143506.GT67687@mdounin.ru>

Hello!

On Wed, Jan 25, 2012 at 03:49:40AM -0500, evilsocket wrote:

> Hi guys, i'm new to this forum (already introduced myself ^^).
> 
> I'm writing a nginx handler module and i need to load a file in a server
> directive inside the config, let's assume it's something like:

[...]

See nginx-devel@ for the reply.  Cross-posting isn't really good 
idea.

Maxim Dounin


From nginx-forum at nginx.us  Wed Jan 25 14:42:02 2012
From: nginx-forum at nginx.us (evilsocket)
Date: Wed, 25 Jan 2012 09:42:02 -0500
Subject: Module API and srv structure initialization.
In-Reply-To: <20120125143506.GT67687@mdounin.ru>
References: <20120125143506.GT67687@mdounin.ru>
Message-ID: <4ee829e07b561ec89fc52060ec911a99.NginxMailingListEnglish@forum.nginx.org>

I guess i'm gonna load it as a main configuration inside the post
configuration handler, thanks so much :)

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221610,221637#msg-221637


From nginx-forum at nginx.us  Wed Jan 25 14:50:47 2012
From: nginx-forum at nginx.us (lockev3.0)
Date: Wed, 25 Jan 2012 09:50:47 -0500
Subject: Default_server catch all block not working
In-Reply-To: <87vco0vpz6.wl%appa@perusio.net>
References: <87vco0vpz6.wl%appa@perusio.net>
Message-ID: <8685c4f3e7a9b75a055b4057913992e7.NginxMailingListEnglish@forum.nginx.org>

Ant?nio P. P. Almeida Wrote:
-------------------------------------------------------
> On 24 Jan 2012 17h57 WET, nginx-forum at nginx.us
> wrote:
> 
> > glubs !!! header host looks like "the missing
> culprit", I'll check
> > it out to confirm your nice guess Mit Rowe
> 
> 444 closes the connection. Doesn't return any 503
> code.

I know i know ...it's commented out


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

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221572,221639#msg-221639


From nginx-forum at nginx.us  Wed Jan 25 15:39:46 2012
From: nginx-forum at nginx.us (lockev3.0)
Date: Wed, 25 Jan 2012 10:39:46 -0500
Subject: Default_server catch all block not working
In-Reply-To: <8685c4f3e7a9b75a055b4057913992e7.NginxMailingListEnglish@forum.nginx.org>
References: <87vco0vpz6.wl%appa@perusio.net>
	<8685c4f3e7a9b75a055b4057913992e7.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <dbd619cbf3be7c8c3d7ea9f3e7e755ee.NginxMailingListEnglish@forum.nginx.org>

Ok as I started thinking later ....what for do I need proxy_set_header
if my intention is to not proxy to Apache any request fall onto this
catch all default_server  ?

To see my config let's say it is as follow (not all parameters to let it
be clearer):

======================    /etc/nginx/sites-enabled/mydomain
server {
        
        listen   80;                
        server_name mydomain.com www.mydomain.com sub1.mydomain.com
sub2.mydomain.com sub3.mydomain.com
        error_page  502 503 504 400 /50x.html;
        error_page  404  /404.php;

        large_client_header_buffers     4 4k;       

        location = /50x.html {
                root /var/www/mydomain/static/;
        }
        location ~* \.(jpg|jpeg|gif|png|css|bmp|ico|txt|html|swf)$ {   

                root /var/www/mydomain/
                expires 1y;
        }
          location / {
                proxy_pass         http://127.0.0.1:60080/;
                proxy_redirect     off;
             ........
          }      
}

So my goal is to define a new vhost so that all requests reaching my
nginx which don't fit any of my server names (all subdomains) are
responded as one of 2 ::  404 returned or forwarding to another external
URL (...whatever it fits better for SEO)

======================    /etc/nginx/sites-enabled/mydomain
server {
       listen 80 default_server;
        server_name  _;
        error_log       /var/log/nginx/000default-error.log error;
       location / {
               proxy_set_header Host $host;
               #return 404;
               rewrite .* http://doesnotexist.com permanent;
       }
}


I must say I tested (sure badly i did ) the proxy_set_header (not sure
what for in my case) and still the same behaviour .....all my clients
away :(

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221572,221640#msg-221640


From nginx-forum at nginx.us  Wed Jan 25 15:55:48 2012
From: nginx-forum at nginx.us (lockev3.0)
Date: Wed, 25 Jan 2012 10:55:48 -0500
Subject: Running other vhosts beside the catc-all default vhost
In-Reply-To: <AANLkTinKnv4_P9mpN6N1PO8-_p-BKguvbBMRUvkqv6zU@mail.gmail.com>
References: <AANLkTinKnv4_P9mpN6N1PO8-_p-BKguvbBMRUvkqv6zU@mail.gmail.com>
Message-ID: <8a4adb62c70e6419504f3af6f2beebd2.NginxMailingListEnglish@forum.nginx.org>

a bit late but, could you tell me how? quite similar situation I have
(see my post at
http://forum.nginx.org/read.php?2,221572,221640#msg-221640) but
desperately failing with all tested combinations !
please Nedim help

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,99440,221641#msg-221641


From appa at perusio.net  Wed Jan 25 16:16:30 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Wed, 25 Jan 2012 16:16:30 +0000
Subject: Default_server catch all block not working
In-Reply-To: <dbd619cbf3be7c8c3d7ea9f3e7e755ee.NginxMailingListEnglish@forum.nginx.org>
References: <87vco0vpz6.wl%appa@perusio.net>
	<8685c4f3e7a9b75a055b4057913992e7.NginxMailingListEnglish@forum.nginx.org>
	<dbd619cbf3be7c8c3d7ea9f3e7e755ee.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <87wr8frc4h.wl%appa@perusio.net>

On 25 Jan 2012 15h39 WET, nginx-forum at nginx.us wrote:

> Ok as I started thinking later ....what for do I need
> proxy_set_header if my intention is to not proxy to Apache any
> request fall onto this catch all default_server ?
>
> To see my config let's say it is as follow (not all parameters to
> let it be clearer):
>
> ======================    /etc/nginx/sites-enabled/mydomain
> server {
>
> listen   80;                
> server_name mydomain.com www.mydomain.com sub1.mydomain.com
> sub2.mydomain.com sub3.mydomain.com
> error_page  502 503 504 400 /50x.html;
> error_page  404  /404.php;
>
> large_client_header_buffers     4 4k;       
>
> location = /50x.html {
> root /var/www/mydomain/static/;
> }
> location ~* \.(jpg|jpeg|gif|png|css|bmp|ico|txt|html|swf)$ {   
>
> root /var/www/mydomain/
> expires 1y;
> }
> location / {
> proxy_pass         http://127.0.0.1:60080/;
> proxy_redirect     off;
> ........
> }      
> }
>
> So my goal is to define a new vhost so that all requests reaching my
> nginx which don't fit any of my server names (all subdomains) are
> responded as one of 2 :: 404 returned or forwarding to another
> external URL (...whatever it fits better for SEO)
>
> ======================    /etc/nginx/sites-enabled/mydomain
> server {
> listen 80 default_server;
> server_name  _;
> error_log       /var/log/nginx/000default-error.log error;
> location / {
> proxy_set_header Host $host;
> #return 404;
> rewrite .* http://doesnotexist.com permanent;
> }
> }

This can be made much simpler:

server {
    listen 80 default_server;
    server_name _;

    error_log /var/log/nginx/000default-error.log error;   

    return 301 http://doesnotexist.com;
}

This redirects to a new host.

For returning a 404 just do:

return 404;

I think that 444 is preferable. No business trying to access a host
that doesn't exist. But it's up to you.

Enable the debug log and set:

proxy_intercept_errors on;

Trace the requests and see how are the 503s generated.

http://nginx.org/en/docs/debugging_log.html

--- appa


From nginx-forum at nginx.us  Wed Jan 25 23:14:43 2012
From: nginx-forum at nginx.us (ggrensteiner)
Date: Wed, 25 Jan 2012 18:14:43 -0500
Subject: Nginx as Load Balancer Connection Issues
In-Reply-To: <a812381f5ffe56a06d2b61547b2e1b0c.NginxMailingListEnglish@forum.nginx.org>
References: <CABYiri-Gwpm2m7pdgLSp0rWa1KxAipTTDH9Ca_xvU3_cszq-6Q@mail.gmail.com>
	<a812381f5ffe56a06d2b61547b2e1b0c.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <6ecc7d03b64833b6f692b7a09d9a4c74.NginxMailingListEnglish@forum.nginx.org>

Have you tried using HTTP 1.1 keepalive connections from nginx to
apache?  They became available in 1.1.4 and will re-use sockets rather
then close them and leaving them in TIME_WAIT

Be sure to remember to turn on keepalive in your apache config as well.

http://nginx.org/en/docs/http/ngx_http_upstream_module.html

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220894,221646#msg-221646


From rami.essaid at gmail.com  Wed Jan 25 23:21:54 2012
From: rami.essaid at gmail.com (Rami Essaid)
Date: Wed, 25 Jan 2012 17:21:54 -0600
Subject: Nginx as Load Balancer Connection Issues
In-Reply-To: <6ecc7d03b64833b6f692b7a09d9a4c74.NginxMailingListEnglish@forum.nginx.org>
References: <CABYiri-Gwpm2m7pdgLSp0rWa1KxAipTTDH9Ca_xvU3_cszq-6Q@mail.gmail.com>
	<a812381f5ffe56a06d2b61547b2e1b0c.NginxMailingListEnglish@forum.nginx.org>
	<6ecc7d03b64833b6f692b7a09d9a4c74.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <7F55D3941DC042E3ABD2AF99C723E3AA@gmail.com>

Out of curiosity why would it keep it in TIME_WAIT if it is closing the connection? 


On Wednesday, January 25, 2012 at 5:14 PM, ggrensteiner wrote:

> Have you tried using HTTP 1.1 keepalive connections from nginx to
> apache? They became available in 1.1.4 and will re-use sockets rather
> then close them and leaving them in TIME_WAIT
> 
> Be sure to remember to turn on keepalive in your apache config as well.
> 
> http://nginx.org/en/docs/http/ngx_http_upstream_module.html
> 
> Posted at Nginx Forum: http://forum.nginx.org/read.php?2,220894,221646#msg-221646
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org (mailto:nginx at nginx.org)
> http://mailman.nginx.org/mailman/listinfo/nginx
> 
> 


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120125/b7a94afb/attachment.html>

From emakyol at gmail.com  Thu Jan 26 09:27:13 2012
From: emakyol at gmail.com (Engin Akyol)
Date: Thu, 26 Jan 2012 03:27:13 -0600
Subject: Passing variables to perl modules in nginx.conf:
Message-ID: <CAOVhSvsZJNppsmfGhyjKwD9krdv-GguUwoCkrM1zFM=C7tE1jA@mail.gmail.com>

Hello,

Is it possible to pass variables to embedded perl modules via
nginx.conf and if so how would that be done?  For instance(approx):


http {
  perl_modules  perl/lib;
  perl_require  My::super_task.pm;
}

Server{
     location @important_task{

                perl My::super_task "A";
        }
}

-Thanks

/e


From zzz at zzz.org.ua  Thu Jan 26 13:33:01 2012
From: zzz at zzz.org.ua (Alexandr Gomoliako)
Date: Thu, 26 Jan 2012 15:33:01 +0200
Subject: Passing variables to perl modules in nginx.conf:
In-Reply-To: <CAOVhSvsZJNppsmfGhyjKwD9krdv-GguUwoCkrM1zFM=C7tE1jA@mail.gmail.com>
References: <CAOVhSvsZJNppsmfGhyjKwD9krdv-GguUwoCkrM1zFM=C7tE1jA@mail.gmail.com>
Message-ID: <CAKb9G9jN9XXvpc0JXzOb9uobeGc357-FFLqShSdA-8vyQN6tYw@mail.gmail.com>

On 1/26/12, Engin Akyol <emakyol at gmail.com> wrote:

> Is it possible to pass variables to embedded perl modules via
> nginx.conf and if so how would that be done?

    set $foo "bar";
    perl My::super_task "A";

In perl:

    sub handler {
        my $r = shift;
        my $foo = $r->variable('foo');
        ...
    }


From nginx-forum at nginx.us  Thu Jan 26 15:30:56 2012
From: nginx-forum at nginx.us (Xeriar)
Date: Thu, 26 Jan 2012 10:30:56 -0500
Subject: Nginx ridiculously slow with PHP+SSL combined
Message-ID: <6622a9ceedd83095bbc4da184bfaf0c9.NginxMailingListEnglish@forum.nginx.org>

So, I'm having an issue with SSL and fastcgi_pass,
specifically to PHP. 

PHP over standard http gets me over 6,000 requests
per second, without stressing the CPU. It's limited
by somaxconn and maximum php children.

Static SSL also easily passes 6,000 requests per
second without stressing the CPU. It's limited
by raw bandwidth. Setting up a proxy on localhost
and forcing it through proxy_pass did not significantly
slow this test down.

Calling the same php file over SSL, however, causes
nginx to fully utilize all cores - php takes nothing -
and only permits 400 requests/second. Taking CPU
usage into account, this is roughly fifty times slower
than the above two tests. Disabling gzip provides
a ~10% boost in speed. Likewise, I got the same
result through proxy_pass.

My server is a quad-core nehalem (E5520 - 2.27 ghz)
with 12 gigs of RAM, running Debian Squeeze.

I've also done these tests on a Wheezy server,
on the hopes that the issue had been resolved
in 1.1.12, but no such luck. I hadn't tweaked
its settings to handle this sort of load, so
the base php and ssl requests/second are 
much less dramatic than the production server,
but the PHP-over-SSL results are the same.

This seems like a rather enormous discrepancy,
but I have no idea where to look. Does anyone have
an idea of what might be going on?
  
Siege reports the same limitation for php over SSL,
but because it doesn't support keepalive it blows
through sockets too fast to meaningfully test
my servers for standard php or ssl calls.

apachebench logs follow.
----

ab -k -c 900 -n 90000 http://amaterasu/test.php
This is ApacheBench, Version 2.3 <$Revision: 655654 $>
Copyright 1996 Adam Twiss, Zeus Technology Ltd,
http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/

Benchmarking amaterasu (be patient)
Completed 9000 requests
Completed 18000 requests
Completed 27000 requests
Completed 36000 requests
Completed 45000 requests
Completed 54000 requests
Completed 63000 requests
Completed 72000 requests
Completed 81000 requests
Completed 90000 requests
Finished 90000 requests


Server Software:        nginx
Server Hostname:        amaterasu
Server Port:            80

Document Path:          /test.php
Document Length:        24 bytes

Concurrency Level:      900
Time taken for tests:   14.159 seconds
Complete requests:      90000
Failed requests:        5754
   (Connect: 0, Receive: 0, Length: 5754, Exceptions: 0)
Write errors:           0
Keep-Alive requests:    0
Total transferred:      14577071 bytes
HTML transferred:       2157071 bytes
Requests per second:    6356.29 [#/sec] (mean)
Time per request:       141.592 [ms] (mean)
Time per request:       0.157 [ms] (mean, across all concurrent
requests)
Transfer rate:          1005.38 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:       58   61   3.4     61     122
Processing:    59   80  83.6     70    3084
Waiting:       59   78  83.4     69    3083
Total:        118  141  83.8    131    3145

Percentage of the requests served within a certain time (ms)
  50%    131
  66%    139
  75%    146
  80%    151
  90%    168
  95%    185
  98%    200
  99%    210
 100%   3145 (longest request)
 
----

ab -k -c 900 -n 90000 https://amaterasu/robots.txt
This is ApacheBench, Version 2.3 <$Revision: 655654 $>
Copyright 1996 Adam Twiss, Zeus Technology Ltd,
http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/

Benchmarking amaterasu (be patient)
Completed 9000 requests
Completed 18000 requests
Completed 27000 requests
Completed 36000 requests
Completed 45000 requests
Completed 54000 requests
Completed 63000 requests
Completed 72000 requests
Completed 81000 requests
Completed 90000 requests
Finished 90000 requests


Server Software:        nginx
Server Hostname:        amaterasu
Server Port:            443
SSL/TLS Protocol:       TLSv1/SSLv3,RC4-SHA,2048,128

Document Path:          /robots.txt
Document Length:        1063 bytes

Concurrency Level:      900
Time taken for tests:   13.838 seconds
Complete requests:      90000
Failed requests:        0
Write errors:           0
Keep-Alive requests:    89708
Total transferred:      118881532 bytes
HTML transferred:       95670000 bytes
Requests per second:    6503.96 [#/sec] (mean)
Time per request:       138.377 [ms] (mean)
Time per request:       0.154 [ms] (mean, across all concurrent
requests)
Transfer rate:          8389.77 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0   20 222.2      0    3959
Processing:    59  117 177.0    107    3496
Waiting:       59  117 177.0    107    3496
Total:         59  137 351.4    107    4083

Percentage of the requests served within a certain time (ms)
  50%    107
  66%    110
  75%    112
  80%    113
  90%    117
  95%    120
  98%    125
  99%   1590
 100%   4083 (longest request)
 
----

ab -k -c 900 -n 9000 https://amaterasu/test.php
This is ApacheBench, Version 2.3 <$Revision: 655654 $>
Copyright 1996 Adam Twiss, Zeus Technology Ltd,
http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/

Benchmarking amaterasu (be patient)
Completed 900 requests
Completed 1800 requests
Completed 2700 requests
Completed 3600 requests
Completed 4500 requests
Completed 5400 requests
Completed 6300 requests
Completed 7200 requests
Completed 8100 requests
Completed 9000 requests
Finished 9000 requests


Server Software:        nginx
Server Hostname:        amaterasu
Server Port:            443
SSL/TLS Protocol:       TLSv1/SSLv3,RC4-SHA,2048,128

Document Path:          /test.php
Document Length:        24 bytes

Concurrency Level:      900
Time taken for tests:   22.271 seconds
Complete requests:      9000
Failed requests:        445
   (Connect: 0, Receive: 0, Length: 445, Exceptions: 0)
Write errors:           0
Keep-Alive requests:    0
Total transferred:      1457620 bytes
HTML transferred:       215620 bytes
Requests per second:    404.11 [#/sec] (mean)
Time per request:       2227.101 [ms] (mean)
Time per request:       2.475 [ms] (mean, across all concurrent
requests)
Transfer rate:          63.92 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:      181 1033 682.6    885    3609
Processing:    59 1169 655.3   1103    3552
Waiting:       59 1157 661.4   1094    3552
Total:        266 2202 796.6   2223    4715

Percentage of the requests served within a certain time (ms)
  50%   2223
  66%   2369
  75%   2474
  80%   2720
  90%   3140
  95%   3775
  98%   3982
  99%   4005
 100%   4715 (longest request)

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221673,221673#msg-221673


From reallfqq-nginx at yahoo.fr  Thu Jan 26 19:16:22 2012
From: reallfqq-nginx at yahoo.fr (B.R.)
Date: Thu, 26 Jan 2012 14:16:22 -0500
Subject: autoindex directive in 'if' section
Message-ID: <CALqce=1zPKy6Npd_PTm-3eePMizcCVDh1Ji0nua9KCV-NWQ+BA@mail.gmail.com>

Hello,

I would like to use the following in my conf file:
location a/ {
 if ($host ~* "") {
  root "";
  autoindex on;
 }
}

But that doesn't work: 'nginx: [emerg] "autoindex" directive is not allowed
here'

What should I do?

Thanks,
---
*B. R.*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120126/566733ec/attachment.html>

From mdounin at mdounin.ru  Thu Jan 26 19:21:19 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Thu, 26 Jan 2012 23:21:19 +0400
Subject: autoindex directive in 'if' section
In-Reply-To: <CALqce=1zPKy6Npd_PTm-3eePMizcCVDh1Ji0nua9KCV-NWQ+BA@mail.gmail.com>
References: <CALqce=1zPKy6Npd_PTm-3eePMizcCVDh1Ji0nua9KCV-NWQ+BA@mail.gmail.com>
Message-ID: <20120126192119.GF67687@mdounin.ru>

Hello!

On Thu, Jan 26, 2012 at 02:16:22PM -0500, B.R. wrote:

> Hello,
> 
> I would like to use the following in my conf file:
> location a/ {
>  if ($host ~* "") {
>   root "";
>   autoindex on;
>  }
> }
> 
> But that doesn't work: 'nginx: [emerg] "autoindex" directive is not allowed
> here'
> 
> What should I do?

http://wiki.nginx.org/IfIsEvil

Maxim Dounin


From emakyol at gmail.com  Thu Jan 26 20:17:42 2012
From: emakyol at gmail.com (Engin Akyol)
Date: Thu, 26 Jan 2012 14:17:42 -0600
Subject: Passing variables to perl modules in nginx.conf:
In-Reply-To: <CAKb9G9jN9XXvpc0JXzOb9uobeGc357-FFLqShSdA-8vyQN6tYw@mail.gmail.com>
References: <CAOVhSvsZJNppsmfGhyjKwD9krdv-GguUwoCkrM1zFM=C7tE1jA@mail.gmail.com>
	<CAKb9G9jN9XXvpc0JXzOb9uobeGc357-FFLqShSdA-8vyQN6tYw@mail.gmail.com>
Message-ID: <CAOVhSvuRj08mFOHDz5n3dszGJcULYm05p6Ao-Ov+JaFT9TJP4Q@mail.gmail.com>

Awesome! That worked.


On Thu, Jan 26, 2012 at 7:33 AM, Alexandr Gomoliako <zzz at zzz.org.ua> wrote:
> On 1/26/12, Engin Akyol <emakyol at gmail.com> wrote:
>
>> Is it possible to pass variables to embedded perl modules via
>> nginx.conf and if so how would that be done?
>
> ? ?set $foo "bar";
> ? ?perl My::super_task "A";
>
> In perl:
>
> ? ?sub handler {
> ? ? ? ?my $r = shift;
> ? ? ? ?my $foo = $r->variable('foo');
> ? ? ? ?...
> ? ?}
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From nginxyz at mail.ru  Thu Jan 26 20:23:04 2012
From: nginxyz at mail.ru (=?UTF-8?B?TWF4?=)
Date: Fri, 27 Jan 2012 00:23:04 +0400
Subject: Nginx ridiculously slow with PHP+SSL combined
In-Reply-To: <6622a9ceedd83095bbc4da184bfaf0c9.NginxMailingListEnglish@forum.nginx.org>
References: <6622a9ceedd83095bbc4da184bfaf0c9.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <E1RqVqa-0007gT-Fa.nginxyz-mail-ru@f250.mail.ru>


26 ?????? 2012, 19:31 ?? "Xeriar" <nginx-forum at nginx.us>:
> 
> This seems like a rather enormous discrepancy,
> but I have no idea where to look. Does anyone have
> an idea of what might be going on?

Take a look at this and you will probably notice
what is going on:
 
> ab -k -c 900 -n 90000 http://amaterasu/test.php
> Complete requests:      90000
> Failed requests:        5754
> Keep-Alive requests:    0
> Transfer rate:          1005.38 [Kbytes/sec] received
>  
> ab -k -c 900 -n 90000 https://amaterasu/robots.txt
> Complete requests:      90000
> Failed requests:        0
> Keep-Alive requests:    89708
> Transfer rate:          8389.77 [Kbytes/sec] received
> 
> ab -k -c 900 -n 9000 https://amaterasu/test.php
> Complete requests:      9000
> Failed requests:        445
> Keep-Alive requests:    0
> Transfer rate:          63.92 [Kbytes/sec] received

The robots.txt via SSL test shows 89708 of 90000
Keep-Alive requests, which means the same SSL
connection was reused 99.67 % of the time, which
is why you got such good test results.

But the test.php via SSL test shows 0 out of 9000
Keep-Alive requests, which means that a new SSL
connection was established for each and every
successful request. Considering how CPU intensive
SSL is, your test results are really not that bad.

Max

From nginx-forum at nginx.us  Fri Jan 27 05:53:18 2012
From: nginx-forum at nginx.us (Xeriar)
Date: Fri, 27 Jan 2012 00:53:18 -0500
Subject: Nginx ridiculously slow with PHP+SSL combined
In-Reply-To: <E1RqVqa-0007gT-Fa.nginxyz-mail-ru@f250.mail.ru>
References: <E1RqVqa-0007gT-Fa.nginxyz-mail-ru@f250.mail.ru>
Message-ID: <30c1e555caf038334a6a62c429bc0367.NginxMailingListEnglish@forum.nginx.org>

I am a moron.

This is embarassing, you'd think that 'siege
reported the same results...doesn't use
keepalive' would have lit a bulb somewhere.

Setting a valid Content-Length gives
a much nicer picture, and it still only uses
half the cpu:

Server Software:        nginx
Server Hostname:        amaterasu
Server Port:            443
SSL/TLS Protocol:       TLSv1/SSLv3,RC4-SHA,2048,128

Document Path:          /test.php
Document Length:        13 bytes

Concurrency Level:      900
Time taken for tests:   13.894 seconds
Complete requests:      90000
Failed requests:        5
   (Connect: 0, Receive: 0, Length: 5, Exceptions: 0)
Write errors:           0
Keep-Alive requests:    89372
Total transferred:      29680613 bytes
HTML transferred:       1169948 bytes
Requests per second:    6477.66 [#/sec] (mean)
Time per request:       138.939 [ms] (mean)
Time per request:       0.154 [ms] (mean, across all concurrent
requests)
Transfer rate:          2086.16 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0   15 146.6      0    2346
Processing:    62  123 110.9    102    3125
Waiting:        0  123 108.5    102    1679
Total:         62  138 237.3    102    3125

Percentage of the requests served within a certain time (ms)
  50%    102
  66%    113
  75%    122
  80%    129
  90%    155
  95%    193
  98%    413
  99%   1204
 100%   3125 (longest request)

----

Thank you for pointing out my stupidity, I'll go hide
in shame now. : )

Max Wrote:
-------------------------------------------------------
> The robots.txt via SSL test shows 89708 of 90000
> Keep-Alive requests, which means the same SSL
> connection was reused 99.67 % of the time, which
> is why you got such good test results.
> 
> But the test.php via SSL test shows 0 out of 9000
> Keep-Alive requests, which means that a new SSL
> connection was established for each and every
> successful request. Considering how CPU intensive
> SSL is, your test results are really not that bad.
> 
> Max
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221673,221683#msg-221683


From daniel.carrillo at gmail.com  Fri Jan 27 11:33:40 2012
From: daniel.carrillo at gmail.com (Daniel Carrillo)
Date: Fri, 27 Jan 2012 12:33:40 +0100
Subject: Possible memory leak nginx 1.1.12 with pcre and jit enabled
In-Reply-To: <201201242317.40871.ne@vbart.ru>
References: <CAK6GuySUx9KmvJ4Ugv9cmVcrHH5Uc2Z=oAPkpQ9Uj+UW9BTjnA@mail.gmail.com>
	<201201242317.40871.ne@vbart.ru>
Message-ID: <CAK6GuyQ_Qk3fDODppsizbfVMv24Q-m-p9JrWFFykxFK3uaZK+Q@mail.gmail.com>

2012/1/24 Valentin V. Bartenev <ne at vbart.ru>:
> On Tuesday 24 January 2012 22:12:07 Daniel Carrillo wrote:
>> Hi all.
>>
>> I compiled nginx against pcre 8.21 and jit enabled. With this scenario
>> in a few days the nginx processes are using more than 100MiB of RAM
>> each. The same scenario with nginx 1.1.12 compiled against the system
>> pcre library and jit disable (not available) the nginx processes are
>> not consuming more than 13MiB each (as expected).
>>
>> CentOS 6.x
>> nginx 1.1.12
>> ./configure --with-pcre=../pcre-8.21/ --with-pcre-jit
>> --with-http_stub_status_module --with-http_ssl_module
>> --without-select_module --without-http_ssi_module
>> --without-http_userid_module --without-poll_module
>> --without-http_ssi_module --without-http_autoindex_module
>> --without-http_geo_module --without-http_map_module
>> --without-http_split_clients_module --without-http_referer_module
>> --without-http_uwsgi_module --without-http_scgi_module
>> --without-http_memcached_module --without-http_limit_conn_module
>> --without-http_limit_req_module --without-http_browser_module
>> --without-http_upstream_ip_hash_module --without-http-cache
>> --without-mail_pop3_module --without-mail_imap_module
>> --without-mail_smtp_module
>>
>> I'm not sure if the memory leak is related to nginx, pcre,
>> configuration or a mix. Please let me know if more information is
>> needed, configuration files, etc.
>>
>
> Yes, please, provide your nginx config.
>
> Which version is your system PCRE library? Does it leak when the "pcre_jit"
> directive is switched off?

Just for update. I sent my configuration privately to Valentin, and he
found a memory leak, he wrote a patch that works as expected, so I
think that will be fixed in 1.1.14.

Thanks so much to Valentin.

Kind regards.


From jkpasanen at gmail.com  Fri Jan 27 14:02:47 2012
From: jkpasanen at gmail.com (Jaakko Pasanen)
Date: Fri, 27 Jan 2012 16:02:47 +0200
Subject: How to proxy_pass POST with headers
Message-ID: <D857F2D8-C9B6-43FD-8AE3-65485E437AA0@gmail.com>

Hi,

I'm using Typo3 as CMS with Apache on Ubuntu 10.04. I'm trying to set Nginx as a reverse proxy with a typo3 extension evo_nginx_boost that uses memcached. I have everything running on the same machine, Nginx on port 80 and apache behind a firewall on port 8085 and memcached behind a on port 11211.

So far I have got everything else working just fine except sending any POST information, ex. logging in, etc. This causes browser to try to fetch images, etc. behind firewall (port 8085) and also Typo3 doesn't know how to parse the url to right place (it's creating nice urls. Instead of foo.bar/index.php?id=123 it uses foo.bar/this/is/the/file.html)

I think that In order to get it work I would somehow need to be able to set proxy_set_header Host $http_host;, but Nginx doesn't let me do that in if-statement (see below). I really don't know how to resolve this and googling around didn't help me, so I would be extremely happy if anybody could give me any feedback! Thanks!

Here is my Nginx conf for the server:

server {
     listen       80; 
     server_name  foo.bar;

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

     location / { 
    
        set $domain "http://foo.bar";
    
     if ($request_method = POST)
        {   
            proxy_pass  http://foo.bar:8085;
            break;
        }   
   
    # set default memcache key 
           set $href $host$request_uri;

      if ($http_cookie ~* "be_typo_user")
        {   
           proxy_pass  http://foo.bar:8085;
           break;
        }   


      if ($http_cookie ~* "nginx_boost_fe_user=([^;]+)(?:;|$)")
        {   
           set $href $host$request_uri$1;
        }   

           set $memcached_key $href;

    # Check if local memcached server can answer this request
           default_type text/html;
           memcached_pass 127.0.0.1:11211;
 
    # Send to app. server if Memcached could not answer the request
          error_page 404 502 = /cache_miss$request_uri;

       }   
    # Main location
      location /cache_miss/
        {   
          proxy_pass http://foo.bar:8085;
          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;
        }   

    location ~ ^/(images|javascript|js|css|flash|media|fileadmin|typo3temp|static|uploads)/  {
          proxy_pass http://foo.bar:8085;
          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;
        }
    }

server {
    listen       80 default_server;
    server_name  _;
    return       301 http://foo.bar$request_uri;
}


---

Jaakko Pasanen



From appa at perusio.net  Fri Jan 27 14:12:43 2012
From: appa at perusio.net (=?UTF-8?B?QW50w7NuaW8=?= P. P. Almeida)
Date: Fri, 27 Jan 2012 14:12:43 +0000
Subject: How to proxy_pass POST with headers
In-Reply-To: <D857F2D8-C9B6-43FD-8AE3-65485E437AA0@gmail.com>
References: <D857F2D8-C9B6-43FD-8AE3-65485E437AA0@gmail.com>
Message-ID: <87sjj1qlno.wl%appa@perusio.net>

On 27 Jan 2012 14h02 WET, jkpasanen at gmail.com wrote:

> Hi,
>
> I'm using Typo3 as CMS with Apache on Ubuntu 10.04. I'm trying to
> set Nginx as a reverse proxy with a typo3 extension evo_nginx_boost
> that uses memcached. I have everything running on the same machine,
> Nginx on port 80 and apache behind a firewall on port 8085 and
> memcached behind a on port 11211.
>
> So far I have got everything else working just fine except sending
> any POST information, ex. logging in, etc. This causes browser to
> try to fetch images, etc. behind firewall (port 8085) and also Typo3
> doesn't know how to parse the url to right place (it's creating nice
> urls. Instead of foo.bar/index.php?id=123 it uses
> foo.bar/this/is/the/file.html)
>
> I think that In order to get it work I would somehow need to be able
> to set proxy_set_header Host $http_host;, but Nginx doesn't let me
> do that in if-statement (see below). I really don't know how to
> resolve this and googling around didn't help me, so I would be
> extremely happy if anybody could give me any feedback! Thanks!

For starters: http://wiki.nginx.org/IfIsEvil

--- appa


From reallfqq-nginx at yahoo.fr  Fri Jan 27 20:43:45 2012
From: reallfqq-nginx at yahoo.fr (reallfqq-nginx at yahoo.fr)
Date: Fri, 27 Jan 2012 20:43:45 +0000 (GMT)
Subject: autoindex directive in 'if' section
In-Reply-To: <20120126192119.GF67687@mdounin.ru>
References: <CALqce=1zPKy6Npd_PTm-3eePMizcCVDh1Ji0nua9KCV-NWQ+BA@mail.gmail.com>
	<20120126192119.GF67687@mdounin.ru>
Message-ID: <1327697025.73119.YahooMailNeo@web24705.mail.ird.yahoo.com>

I know this Wiki page...
I guess return/rewrite are useless here: I only want to change the root and activate the directory listing when the user is on a specific directory of a specific domain.

Any more help ? :\

Thx,
B.R.



________________________________
 De?: Maxim Dounin <mdounin at mdounin.ru>
??: nginx at nginx.org 
Envoy? le : Jeudi 26 janvier 2012 14h21
Objet?: Re: autoindex directive in 'if' section
 
Hello!

On Thu, Jan 26, 2012 at 02:16:22PM -0500, B.R. wrote:

> Hello,
> 
> I would like to use the following in my conf file:
> location a/ {
>? if ($host ~* "") {
>?  root "";
>?  autoindex on;
>? }
> }
> 
> But that doesn't work: 'nginx: [emerg] "autoindex" directive is not allowed
> here'
> 
> What should I do?

http://wiki.nginx.org/IfIsEvil

Maxim Dounin

_______________________________________________
nginx mailing list
nginx at nginx.org
http://mailman.nginx.org/mailman/listinfo/nginx
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120127/10330f4d/attachment.html>

From jkpasanen at gmail.com  Fri Jan 27 21:13:12 2012
From: jkpasanen at gmail.com (Jaakko Pasanen)
Date: Fri, 27 Jan 2012 23:13:12 +0200
Subject: How to proxy_pass POST with headers
In-Reply-To: <87sjj1qlno.wl%appa@perusio.net>
References: <D857F2D8-C9B6-43FD-8AE3-65485E437AA0@gmail.com>
	<87sjj1qlno.wl%appa@perusio.net>
Message-ID: <113DB86D-7DB7-4E09-BC50-5901D70AA4F0@gmail.com>

Ant?nio P. P. Almeida kirjoitti 27.1.2012 kello 16.12:

> For starters: http://wiki.nginx.org/IfIsEvil

This actually solved my problem!

I changed the if statement to:

if ($request_method = POST)
        {
            return 418;
        } 

and added:

error_page 418 = @post;

location @post {
         proxy_pass http://foo.bar:8085;
         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;
    }

Now it works flawlessly!

Thanks!

---

Jaakko Pasanen

From nginxyz at mail.ru  Fri Jan 27 21:27:29 2012
From: nginxyz at mail.ru (=?UTF-8?B?TWF4?=)
Date: Sat, 28 Jan 2012 01:27:29 +0400
Subject: How to proxy_pass POST with headers
In-Reply-To: <D857F2D8-C9B6-43FD-8AE3-65485E437AA0@gmail.com>
References: <D857F2D8-C9B6-43FD-8AE3-65485E437AA0@gmail.com>
Message-ID: <E1RqtKT-0005Rc-Ra.nginxyz-mail-ru@f193.mail.ru>


27 ?????? 2012, 18:03 ?? Jaakko Pasanen <jkpasanen at gmail.com>:

> I think that In order to get it work I would somehow need to be able to set
> proxy_set_header Host $http_host;, but Nginx doesn't let me do that in
> if-statement (see below). I really don't know how to resolve this and googling
> around didn't help me, so I would be extremely happy if anybody could give me
> any feedback! Thanks!

This should work:

location / {
  if ($request_method = POST) {
   rewrite ^/(.*)$ /post_redirect/$1 last;
  }
}

location ~ ^/post_redirect/(.*)$ {
  internal;
  proxy_set_header Host $http_host;
  set $proxy_url http://foo.bar:8085/$1;
  if ($args) {
    set $proxy_url http://foo.bar:8085/$1?$args;
  }
  proxy_pass $proxy_url;
}

Since rewrite is guaranteed to work without side-effects
inside if blocks, the safest approach would be to first
use rewrite inside the if block (with last) to change the
URI to an internal URI and then handle the rest inside
the internal URI location block.

Max

From nginxyz at mail.ru  Fri Jan 27 21:51:49 2012
From: nginxyz at mail.ru (=?UTF-8?B?TWF4?=)
Date: Sat, 28 Jan 2012 01:51:49 +0400
Subject: autoindex directive in 'if' section
In-Reply-To: <1327697025.73119.YahooMailNeo@web24705.mail.ird.yahoo.com>
References: <CALqce=1zPKy6Npd_PTm-3eePMizcCVDh1Ji0nua9KCV-NWQ+BA@mail.gmail.com>
	<20120126192119.GF67687@mdounin.ru>
	<1327697025.73119.YahooMailNeo@web24705.mail.ird.yahoo.com>
Message-ID: <E1Rqti1-00007r-PH.nginxyz-mail-ru@f279.mail.ru>

 28 ?????? 2012, 00:44 ?? "reallfqq-nginx at yahoo.fr" : > I know this Wiki page... > I guess return/rewrite are useless here: I only want to change the root and > activate the directory listing when the user is on a specific directory of a > specific domain. > > Any more help ? :\ I just explained how this is handled here: http://mailman.nginx.org/pipermail/nginx/2012-January/031622.html In your case use the rewrite inside the if block and and then change root and set autoindex inside the internal URI block. Max
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120128/53c1a696/attachment.html>

From reallfqq-nginx at yahoo.fr  Fri Jan 27 22:41:50 2012
From: reallfqq-nginx at yahoo.fr (B.R.)
Date: Fri, 27 Jan 2012 17:41:50 -0500
Subject: autoindex directive in 'if' section
In-Reply-To: <E1Rqti1-00007r-PH.nginxyz-mail-ru@f279.mail.ru>
References: <CALqce=1zPKy6Npd_PTm-3eePMizcCVDh1Ji0nua9KCV-NWQ+BA@mail.gmail.com>
	<20120126192119.GF67687@mdounin.ru>
	<1327697025.73119.YahooMailNeo@web24705.mail.ird.yahoo.com>
	<E1Rqti1-00007r-PH.nginxyz-mail-ru@f279.mail.ru>
Message-ID: <CALqce=0jfPsdE-M7HPyGiiEs-JznPYuXrsp1zcq=MVL+nUbrTg@mail.gmail.com>

Thanks for the help!
Needed some other tricks playing with the path...

Many thanks again! ;o)
---
*B. R.*


On Fri, Jan 27, 2012 at 16:51, Max <nginxyz at mail.ru> wrote:

>  28 ?????? 2012, 00:44 ?? "reallfqq-nginx at yahoo.fr" **: > I know this
> Wiki page... > I guess return/rewrite are useless here: I only want to
> change the root and > activate the directory listing when the user is on a
> specific directory of a > specific domain. > > Any more help ? :\ I just
> explained how this is handled here:
> http://mailman.nginx.org/pipermail/nginx/2012-January/031622.html In your
> case use the rewrite inside the if block and and then change root and set
> autoindex inside the internal URI block. Max
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120127/bdb96687/attachment-0001.html>

From reallfqq-nginx at yahoo.fr  Fri Jan 27 23:02:11 2012
From: reallfqq-nginx at yahoo.fr (B.R.)
Date: Fri, 27 Jan 2012 18:02:11 -0500
Subject: autoindex directive in 'if' section
In-Reply-To: <CALqce=0jfPsdE-M7HPyGiiEs-JznPYuXrsp1zcq=MVL+nUbrTg@mail.gmail.com>
References: <CALqce=1zPKy6Npd_PTm-3eePMizcCVDh1Ji0nua9KCV-NWQ+BA@mail.gmail.com>
	<20120126192119.GF67687@mdounin.ru>
	<1327697025.73119.YahooMailNeo@web24705.mail.ird.yahoo.com>
	<E1Rqti1-00007r-PH.nginxyz-mail-ru@f279.mail.ru>
	<CALqce=0jfPsdE-M7HPyGiiEs-JznPYuXrsp1zcq=MVL+nUbrTg@mail.gmail.com>
Message-ID: <CALqce=2=-SEWPtBHB-A_7bny77fwerG+XiVNCnRMhKTnnu9QjA@mail.gmail.com>

Erk...
Didn't make all my tests :o)

To end the job, I need to get rid of the 'post_rewrite' part of the path,
so my files get served as '/path/to/files' and not
'/post_rewrite/path/to/files'.
I tried rewrite, but of course that gets me out of the current location
block......

Should I replace the content of $uri? Is there any variable regex
substitution mechanism in nginx?
Couldn't find that on Google nor wiki. Maybe I am not searching for the
right thing.
---
*B. R.*


On Fri, Jan 27, 2012 at 17:41, B.R. <reallfqq-nginx at yahoo.fr> wrote:

> Thanks for the help!
> Needed some other tricks playing with the path...
>
> Many thanks again! ;o)
> ---
> *B. R.*
>
>
> On Fri, Jan 27, 2012 at 16:51, Max <nginxyz at mail.ru> wrote:
>
>>  28 ?????? 2012, 00:44 ?? "reallfqq-nginx at yahoo.fr" **: > I know this
>> Wiki page... > I guess return/rewrite are useless here: I only want to
>> change the root and > activate the directory listing when the user is on a
>> specific directory of a > specific domain. > > Any more help ? :\ I just
>> explained how this is handled here:
>> http://mailman.nginx.org/pipermail/nginx/2012-January/031622.html In
>> your case use the rewrite inside the if block and and then change root and
>> set autoindex inside the internal URI block. Max
>>
>> _______________________________________________
>> nginx mailing list
>> nginx at nginx.org
>> http://mailman.nginx.org/mailman/listinfo/nginx
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120127/d795d7a4/attachment.html>

From nginxyz at mail.ru  Sat Jan 28 00:36:47 2012
From: nginxyz at mail.ru (=?UTF-8?B?TWF4?=)
Date: Sat, 28 Jan 2012 04:36:47 +0400
Subject: autoindex directive in 'if' section
In-Reply-To: <CALqce2=-SEWPtBHB-A_7bny77fwerG+XiVNCnRMhKTnnu9QjA@mail.gmail.com>
References: <CALqce=0jfPsdE-M7HPyGiiEs-JznPYuXrsp1zcq=MVL+nUbrTg@mail.gmail.com>
	<CALqce=1zPKy6Npd_PTm-3eePMizcCVDh1Ji0nua9KCV-NWQ+BA@mail.gmail.com>
	<CALqce2=-SEWPtBHB-A_7bny77fwerG+XiVNCnRMhKTnnu9QjA@mail.gmail.com>
Message-ID: <E1RqwHf-0004KC-61.nginxyz-mail-ru@f203.mail.ru>

 28 ?????? 2012, 03:11 ?? "B.R." : > To end the job, I need to get rid of the 'post_rewrite' part of the path, > so my files get served as '/path/to/files' and not > '/post_rewrite/path/to/files'. > I tried rewrite, but of course that gets me out of the current location > block...... > > Should I replace the content of $uri? Is there any variable regex > substitution mechanism in nginx? > Couldn't find that on Google nor wiki. Maybe I am not searching for the > right thing. http://wiki.nginx.org/HttpCoreModule#alias Use alias instead of root, it does exactly what you need: location ~ ^/post_rewrite/(.*)$ { internal; autoindex on; alias /sandbox/$1; } Max
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120128/18598941/attachment.html>

From reallfqq-nginx at yahoo.fr  Sat Jan 28 03:16:06 2012
From: reallfqq-nginx at yahoo.fr (B.R.)
Date: Fri, 27 Jan 2012 22:16:06 -0500
Subject: autoindex directive in 'if' section
In-Reply-To: <E1RqwHf-0004KC-61.nginxyz-mail-ru@f203.mail.ru>
References: <CALqce=0jfPsdE-M7HPyGiiEs-JznPYuXrsp1zcq=MVL+nUbrTg@mail.gmail.com>
	<CALqce=1zPKy6Npd_PTm-3eePMizcCVDh1Ji0nua9KCV-NWQ+BA@mail.gmail.com>
	<CALqce2=-SEWPtBHB-A_7bny77fwerG+XiVNCnRMhKTnnu9QjA@mail.gmail.com>
	<E1RqwHf-0004KC-61.nginxyz-mail-ru@f203.mail.ru>
Message-ID: <CALqce=3R5qzU0Aj3UhDJ4K2geHEa6zh-7Z_2dX5+ayQY6n6HgA@mail.gmail.com>

So simple...
Thanks for everything Max, it works perfectly! ;o)
---
*B. R.*


On Fri, Jan 27, 2012 at 19:36, Max <nginxyz at mail.ru> wrote:

>  28 ?????? 2012, 03:11 ?? "B.R." **: > To end the job, I need to get rid
> of the 'post_rewrite' part of the path, > so my files get served as
> '/path/to/files' and not > '/post_rewrite/path/to/files'. > I tried
> rewrite, but of course that gets me out of the current location >
> block...... > > Should I replace the content of $uri? Is there any variable
> regex > substitution mechanism in nginx? > Couldn't find that on Google nor
> wiki. Maybe I am not searching for the > right thing.
> http://wiki.nginx.org/HttpCoreModule#alias Use alias instead of root, it
> does exactly what you need: location ~ ^/post_rewrite/(.*)$ { internal;
> autoindex on; alias /sandbox/$1; } Max
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120127/ad7137cc/attachment.html>

From ne at vbart.ru  Sat Jan 28 09:59:39 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Sat, 28 Jan 2012 13:59:39 +0400
Subject: How to proxy_pass POST with headers
In-Reply-To: <E1RqtKT-0005Rc-Ra.nginxyz-mail-ru@f193.mail.ru>
References: <D857F2D8-C9B6-43FD-8AE3-65485E437AA0@gmail.com>
	<E1RqtKT-0005Rc-Ra.nginxyz-mail-ru@f193.mail.ru>
Message-ID: <201201281359.40024.ne@vbart.ru>

On Saturday 28 January 2012 01:27:29 Max wrote:
[...]
> location ~ ^/post_redirect/(.*)$ {
>   internal;
>   proxy_set_header Host $http_host;
>   set $proxy_url http://foo.bar:8085/$1;
>   if ($args) {
>     set $proxy_url http://foo.bar:8085/$1?$args;
>   }
>   proxy_pass $proxy_url;
> }
[...]

  proxy_pass http://foo.bar:8085/$1$is_args$args;

 but, actually, this would be better:

  location /post_redirect/ {
      internal;
      proxy_pass http://foo.bar:8085/;
      proxy_redirect off;
  }

 wbr, Valentin V. Bartenev


From nginx-forum at nginx.us  Sat Jan 28 10:52:38 2012
From: nginx-forum at nginx.us (skyroski)
Date: Sat, 28 Jan 2012 05:52:38 -0500
Subject: client_max_body_size per $request_uri
Message-ID: <ba84157e8c18b6918339a80d72e40c29.NginxMailingListEnglish@forum.nginx.org>

Hi, I have found similar topics on discussion with location based
request but because I'm using a PHP software that by default handles
everything through a single PHP file (including processing POST uploads
of files) I would like to be able to use 

if ($request_uri ~* upload$) {
    client_max_body_size 200M;
}

for example, so that all URI with the word upload at the end will
automatically get set a larger client_max_body_size.

But given that the above syntax is currently impossible. How would I be
able to achieve the same result?

The way my nginx conf works is via:

upstream php {
   server unix:/php.sock;
}

if (!-e $request_filename) {
  rewrite ^/(.*)$ /index.php/$1 last;
  break;
}

location ~ \.php$ {
  fastcgi_pass php;
  include fastcgi_params;
}

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221727,221727#msg-221727


From edho at myconan.net  Sat Jan 28 11:35:47 2012
From: edho at myconan.net (Edho Arief)
Date: Sat, 28 Jan 2012 18:35:47 +0700
Subject: client_max_body_size per $request_uri
In-Reply-To: <ba84157e8c18b6918339a80d72e40c29.NginxMailingListEnglish@forum.nginx.org>
References: <ba84157e8c18b6918339a80d72e40c29.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CABTjkKnjmMGMadOy1Mk_OeynqjLu6h0STXZjD-oKG5vepTviPg@mail.gmail.com>

On Sat, Jan 28, 2012 at 5:52 PM, skyroski <nginx-forum at nginx.us> wrote:
> Hi, I have found similar topics on discussion with location based
> request but because I'm using a PHP software that by default handles
> everything through a single PHP file (including processing POST uploads
> of files) I would like to be able to use
>
> if ($request_uri ~* upload$) {
> ? ?client_max_body_size 200M;
> }

My guess says map would work.

http://wiki.nginx.org/HttpMapModule

-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From ne at vbart.ru  Sat Jan 28 11:57:19 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Sat, 28 Jan 2012 15:57:19 +0400
Subject: client_max_body_size per $request_uri
In-Reply-To: <ba84157e8c18b6918339a80d72e40c29.NginxMailingListEnglish@forum.nginx.org>
References: <ba84157e8c18b6918339a80d72e40c29.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <201201281557.19987.ne@vbart.ru>

On Saturday 28 January 2012 14:52:38 skyroski wrote:
> Hi, I have found similar topics on discussion with location based
> request but because I'm using a PHP software that by default handles
> everything through a single PHP file (including processing POST uploads
> of files) I would like to be able to use
> 
> if ($request_uri ~* upload$) {
>     client_max_body_size 200M;
> }
> 

If you want something like "if ($uri ..." then you should use the "location"
directive: http://www.nginx.org/en/docs/http/ngx_http_core_module.html#location

btw, http://wiki.nginx.org/IfIsEvil

[...]
> if (!-e $request_filename) {
>   rewrite ^/(.*)$ /index.php/$1 last;
>   break;
> }

Use try_files instead:
http://www.nginx.org/en/docs/http/ngx_http_core_module.html#try_files

e.g.:

    location / {
        try_files $uri @index_php;
    }

    location ~* upload$ {
        client_max_body_size 200M;
        ...
    }

    location ~ \.php$ {
        fastcgi_pass php;
        include fastcgi_params;
    }

    location @index_php {
        fastcgi_pass php;
        include fastcgi_params;
        fastcgi_param PATH_INFO   $uri;
        fastcgi_param SCRIPT_NAME /path/to/your/index.php;
    }

 wbr, Valentin V. Bartenev


From nginx-forum at nginx.us  Sat Jan 28 13:16:57 2012
From: nginx-forum at nginx.us (skyroski)
Date: Sat, 28 Jan 2012 08:16:57 -0500
Subject: client_max_body_size per $request_uri
In-Reply-To: <201201281557.19987.ne@vbart.ru>
References: <201201281557.19987.ne@vbart.ru>
Message-ID: <a0cc71709a62798222d7afc51eea4164.NginxMailingListEnglish@forum.nginx.org>

For the record, I used a combination of a similar issue in the mailing
list and a combination of what @Valentin suggested.

In the end, this is what works for me:

if ($request_uri ~* upload$) {
  rewrite ^/(.*)$ /upload/$1 last;
}

location ~ /upload/(.*)$ {
  internal;
  client_max_body_size 200M;
  fastcgi_param PHP_VALUE ....;
  fastcgi_pass php;
  fastcgi_param SCRIPT_FILENAME $document_root/index.php;
  include fastcgi_params;
  fastcgi_param QUERY_STRING /$1;
  fastcgi_param DOCUMENT_URI /index.php;
  fastcgi_param SCRIPT_FILENAME /index.php;
}

All the rest I didn't have to change with the exception of adding an
extra clause to the location \.php$ block to restore PHP_VALUE to
default settings. 

I understand around-abouts why IF is evil but if it's supposed to be
safe for rewrites and return then I'd rather take advantage of that
instead of using location / with try files so that I can deal with query
strings attachments easier and it ends up shorter to implement.

@Valentin's solution would have meant I had to maintain 3 different
fastcgi_pass php location blocks as one would also be required for
location ~ upload$ {. 

Thanks for all the hints and tips.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221727,221730#msg-221730


From edho at myconan.net  Sat Jan 28 13:40:39 2012
From: edho at myconan.net (Edho Arief)
Date: Sat, 28 Jan 2012 20:40:39 +0700
Subject: client_max_body_size per $request_uri
In-Reply-To: <a0cc71709a62798222d7afc51eea4164.NginxMailingListEnglish@forum.nginx.org>
References: <201201281557.19987.ne@vbart.ru>
	<a0cc71709a62798222d7afc51eea4164.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CABTjkKmnt+ZMvHD+51YDqRuuH93ezEY1gLRX_NS7xEPKyrJk7Q@mail.gmail.com>

On Sat, Jan 28, 2012 at 8:16 PM, skyroski <nginx-forum at nginx.us> wrote:
> For the record, I used a combination of a similar issue in the mailing
> list and a combination of what @Valentin suggested.
>
> In the end, this is what works for me:
>
> if ($request_uri ~* upload$) {
> ?rewrite ^/(.*)$ /upload/$1 last;
> }
>
> location ~ /upload/(.*)$ {
> ?internal;
> ?client_max_body_size 200M;
> ?fastcgi_param PHP_VALUE ....;
> ?fastcgi_pass php;
> ?fastcgi_param SCRIPT_FILENAME $document_root/index.php;
> ?include fastcgi_params;
> ?fastcgi_param QUERY_STRING /$1;
> ?fastcgi_param DOCUMENT_URI /index.php;
> ?fastcgi_param SCRIPT_FILENAME /index.php;
> }
>
> All the rest I didn't have to change with the exception of adding an
> extra clause to the location \.php$ block to restore PHP_VALUE to
> default settings.
>
> I understand around-abouts why IF is evil but if it's supposed to be
> safe for rewrites and return then I'd rather take advantage of that
> instead of using location / with try files so that I can deal with query
> strings attachments easier and it ends up shorter to implement.
>

too bad this doesn't work :/

map $request_uri $uplimit {
    ~upload$ 200m;
    default 1m;
}
client_max_body_size $uplimit;

-- 
O< ascii ribbon campaign - stop html mail - www.asciiribbon.org


From nginx-forum at nginx.us  Sat Jan 28 14:19:08 2012
From: nginx-forum at nginx.us (waydenie)
Date: Sat, 28 Jan 2012 09:19:08 -0500
Subject: imaps proxy with imaps backend?
Message-ID: <adcbd776dc53055b64de07751cd0e7f3.NginxMailingListEnglish@forum.nginx.org>

Hello,

>From the forum posts that I can find, imaps/pops is not supported on the
backend of an nginx proxy, is this true? Is this a planned feature?

I ask because we have a situation where we would like to use our common
authentication service with an outside service provider, however we will
not be allowed to send any of our passwords to them (ie we can't sync
our authentication databases).

We were looking at using nginx to provide imaps to our users and
authenticate them through our db, then proxy them to the outside
provider using a stored 'provider password' for that user (ie. the user
connects to us via imaps (or pops), we authenticate them against our
common authentication service, then look up their provider password
(just a random string, which they dont know) and complete the backend
proxy connection over imaps.

In the absence of imaps on the backend side of an nginx proxy, does
anyone have any other suggestions for setting this up?

Thanks,
Wayde.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221732,221732#msg-221732


From gelonida at gmail.com  Sat Jan 28 15:41:50 2012
From: gelonida at gmail.com (Gelonida N)
Date: Sat, 28 Jan 2012 16:41:50 +0100
Subject: nginx doesn't seem to log all accesses forwarded to usgi
Message-ID: <jg14vv$lil$1@dough.gmane.org>

Hi everybody,
I hope the message does NOT show up three times.
My news client always reports errors and I don't know whether the
message was sent out or not.


I'm having following setup:

nginx <-unix_socket-> uwsgi -> django

What I noticed is, that some requests of the uwsgi.log file do NOT show
up in nginx' access_log file.


Example:
in uwsgi.log I can see:
[pid: 1205|app: 0|req: 537/1453] xx.xx.xx.33 () {40 vars in 1794 bytes}
[Thu Jan 26 18:42:31 2012] POST /my_url/run?var=123 => generated 129
bytes in 94 msecs (HTTP/1.0 200) 3 headers in 174 bytes (1 switches on
core 0)

However I don't see any corresponding request logged in nginx log.

Is this the expected behaviour?
What could be the reason?

Ideally I would like to see all accesses in nginx' access file.





My nginx.conf
#--------------
worker_processes  2;
events {
    worker_connections  1024;
}
http {
    include       mime.types;
    default_type  application/octet-stream;

    sendfile        on;
    keepalive_timeout  65;

    include sites-enabled/*;

    uwsgi_temp_path  /etc/uwsgi/;

}

# sites_enabled/ssl;
# ---------------------------------------
The probably relevant nginx config lines:
upstream django {
    ip_hash;
    server unix:/var/sock/nginx.sock;
}


server {
    listen  443;
    server_name  myserver;
    access_log  /var/log/nginx/access.log;
    error_log  /var/log/nginx/error.log;

    ssl                  on;
    # cert info . . . .
    ssl_verify_client optional;
    ssl_session_timeout  5m;
    # cipher info

    location ^~ /static {
        root   /my/static;
        index  index.html index.htm;
    }

    location ^~ /registration {
        uwsgi_pass django;
        uwsgi_param UWSGI_SCHEME https;
        include uwsgi_params;
    }

}


From gelonida at gmail.com  Sat Jan 28 15:13:19 2012
From: gelonida at gmail.com (Gelonida N)
Date: Sat, 28 Jan 2012 16:13:19 +0100
Subject: some requests no tshowing up in access_log
Message-ID: <jg13ag$c7f$1@dough.gmane.org>

Hi everybody,

I'm having following setup:

nginx <-unix_socket-> uwsgi -> django

What I noticed is, that some accesses which are successfully forwarded
to uwsgi are NOT logged in my nginx access_log file.

Is this the expected behaviour?
Ideally I would like, that every access is logged by nginx, sucnh, that
I can quickly check how many and which requests were hitting / going
through nginx.


Example:
in uwsgi.log I can see:
[pid: 1205|app: 0|req: 537/1453] xx.xx.xx.33 () {40 vars in 1794 bytes}
[Thu Jan 26 18:42:31 2012] POST /my_url/run?var=123 => generated 129
bytes in 94 msecs (HTTP/1.0 200) 3 headers in 174 bytes (1 switches on
core 0)

However I don't see any corresponding request in nginx log.

This is driving me rather crazy while trying to debug some issues.

What could be the reason?


My nginx.conf
#--------------
worker_processes  2;
events {
    worker_connections  1024;
}
http {
    include       mime.types;
    default_type  application/octet-stream;

    sendfile        on;
    keepalive_timeout  65;

    include sites-enabled/*;

    uwsgi_temp_path  /etc/uwsgi/;

}

# sites_enabled/ssl;
# ---------------------------------------
The probably relevant nginx config lines:
upstream django {
    ip_hash;
    server unix:/var/sock/nginx.sock;
}


server {
    listen  443;
    server_name  myserver;
    access_log  /var/log/nginx/access.log;
    error_log  /var/log/nginx/error.log;

    ssl                  on;
    # cert info . . . .
    ssl_verify_client optional;
    ssl_session_timeout  5m;
    # cipher info

    location ^~ /static {
        root   /my/static;
        index  index.html index.htm;
    }

    location ^~ /registration {
        uwsgi_pass django;
        uwsgi_param UWSGI_SCHEME https;
        include uwsgi_params;
    }

}


From gelonida at gmail.com  Sat Jan 28 15:38:46 2012
From: gelonida at gmail.com (Gelonida)
Date: Sat, 28 Jan 2012 16:38:46 +0100
Subject: nginx doesn't seem to log all accesses (some uwsgi accesses are
	missing)
Message-ID: <jg14qd$jv1$1@dough.gmane.org>

Hi everybody,
I hope the message does NOT show up three times.
My news client always reports errors and I don't know whether the 
message was sent out or not.


I'm having following setup:

nginx <-unix_socket-> uwsgi -> django

What I noticed is, that some requests of the uwsgi.log file do NOT show 
up in nginx' access_log file.


Example:
in uwsgi.log I can see:
[pid: 1205|app: 0|req: 537/1453] xx.xx.xx.33 () {40 vars in 1794 bytes} 
[Thu Jan 26 18:42:31 2012] POST /my_url/run?var=123 => generated 129 
bytes in 94 msecs (HTTP/1.0 200) 3 headers in 174 bytes (1 switches on 
core 0)

However I don't see any corresponding request logged in nginx log.

Is this the expected behaviour?
What could be the reason?

Ideally I would like to see all accesses in nginx' access file.





My nginx.conf
#--------------
worker_processes  2;
events {
     worker_connections  1024;
}
http {
     include       mime.types;
     default_type  application/octet-stream;

     sendfile        on;
     keepalive_timeout  65;

     include sites-enabled/*;

     uwsgi_temp_path  /etc/uwsgi/;

}

# sites_enabled/ssl;
# ---------------------------------------
The probably relevant nginx config lines:
upstream django {
     ip_hash;
     server unix:/var/sock/nginx.sock;
}


server {
     listen  443;
     server_name  myserver;
     access_log  /var/log/nginx/access.log;
     error_log  /var/log/nginx/error.log;

     ssl                  on;
     # cert info . . . .
     ssl_verify_client optional;
     ssl_session_timeout  5m;
     # cipher info

     location ^~ /static {
         root   /my/static;
         index  index.html index.htm;
     }

     location ^~ /registration {
         uwsgi_pass django;
         uwsgi_param UWSGI_SCHEME https;
         include uwsgi_params;
     }

}


From nginx-forum at nginx.us  Sun Jan 29 10:37:57 2012
From: nginx-forum at nginx.us (trojan2748)
Date: Sun, 29 Jan 2012 05:37:57 -0500
Subject: Can i collect stats without modifying nginx core code?
Message-ID: <6f9b393ce44cc2a0241e70a0fbd01017.NginxMailingListEnglish@forum.nginx.org>

Hello,
  I'm trying to write a module that will collect extended stats, that is
more then what is offered by stub_status. I've dumped a couple of stats,
but had to modify source code outside the module. Modifying vanilla
source isn't an option for me. So before I begin, I was wondering if I
could get some information:

1) Can I collect stats on every request? I want to know how many bytes
were written in the body and headers out to the client.
2) Is it possible to collect stats on reverse proxy's? IE: how many
upstream connections are there, have there been, failed, 50x's?
3) Do this all inside of a module?

Thanks in advance.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221739,221739#msg-221739


From calin at presslabs.com  Sun Jan 29 17:15:47 2012
From: calin at presslabs.com (Calin Don)
Date: Sun, 29 Jan 2012 19:15:47 +0200
Subject: Invalidate cache entry using X-Accel-Expires
Message-ID: <CAEOe2JyK8sTZc-R0sjL6vhPNf629uYaWiQ4qs56=dZYUKvHmVQ@mail.gmail.com>

Hello,

Right now I'm invalidating cache entries using X-Accel-Expires 1 response
with 302 redirect of the resource to itself. I want to ask for your opinion
on two issues:

1. Do you see other response code (other than 302) more suitable for this
action?
2. I think that X-Accel-Expires 0 should also invalidate the cache entry,
beside not caching item at all, since it says this resource should expire
in 0 seconds. What's your opinion on this?

Thanks!
Calin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120129/7c17d7e2/attachment.html>

From mdounin at mdounin.ru  Sun Jan 29 23:03:53 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 30 Jan 2012 03:03:53 +0400
Subject: nginx doesn't seem to log all accesses (some uwsgi accesses are
	missing)
In-Reply-To: <jg14qd$jv1$1@dough.gmane.org>
References: <jg14qd$jv1$1@dough.gmane.org>
Message-ID: <20120129230352.GO67687@mdounin.ru>

Hello!

On Sat, Jan 28, 2012 at 04:38:46PM +0100, Gelonida wrote:

> Hi everybody,
> I hope the message does NOT show up three times.
> My news client always reports errors and I don't know whether the
> message was sent out or not.

It was shown up at least 2 times.  It may be not a good idea to 
post via gmane, probably subscribing directly instead would be 
better.

> I'm having following setup:
> 
> nginx <-unix_socket-> uwsgi -> django
> 
> What I noticed is, that some requests of the uwsgi.log file do NOT
> show up in nginx' access_log file.
> 
> 
> Example:
> in uwsgi.log I can see:
> [pid: 1205|app: 0|req: 537/1453] xx.xx.xx.33 () {40 vars in 1794
> bytes} [Thu Jan 26 18:42:31 2012] POST /my_url/run?var=123 =>
> generated 129 bytes in 94 msecs (HTTP/1.0 200) 3 headers in 174
> bytes (1 switches on core 0)
> 
> However I don't see any corresponding request logged in nginx log.
> 
> Is this the expected behaviour?
> What could be the reason?
> 
> Ideally I would like to see all accesses in nginx' access file.

All accesses are expected to be logged, unless logging is switched 
off.  Note though:

1) requests are logged on completion, and if request processing 
and/or sending response to client takes a while - log line won't 
appear till completion

2) requests are logged in configuration of a location where 
request ends; this may be quite different from one where 
processing reaches backend, especially if X-Accel-Redirect or 
error_page redirections are used.

3) one request may hit easily hit backends several times (see 
above about redirections)

If you think something is wrong, you may build debug log to see 
what actually happens, see 
http://nginx.org/en/docs/debugging_log.html.

Note that you have to configure one at global level, and make sure 
it's not overriten at more specific levels (best solution is to 
comment out all other error_log directives), else you are risking 
to miss debug log data as well.

> My nginx.conf
> #--------------
> worker_processes  2;
> events {
>     worker_connections  1024;
> }
> http {
>     include       mime.types;
>     default_type  application/octet-stream;
> 
>     sendfile        on;
>     keepalive_timeout  65;
> 
>     include sites-enabled/*;
> 
>     uwsgi_temp_path  /etc/uwsgi/;
> 
> }
> 
> # sites_enabled/ssl;
> # ---------------------------------------
> The probably relevant nginx config lines:

Actually, full configuration is relevant, including all 
other includes.  Else the information is mostly useless.

[...]

Maxim Dounin


From jombik at platon.org  Mon Jan 30 03:09:33 2012
From: jombik at platon.org (Ondrej Jombik)
Date: Mon, 30 Jan 2012 04:09:33 +0100 (CET)
Subject: MP4 (H264) migrating from lighttpd to nginx
Message-ID: <alpine.DEB.2.00.1201300400040.29347@athena.platon.sk>

It is great that MP4 module came out. So we decided to migrate our video
streaming servers to nginx. In fact those are our last non-nginx
webservers (not counting IIS).

However we have some troubles. In our Flash player, MP4 video starts
play only after it is completelly loaded. I thought there is problem
with player, so I tried open-source flowplayer and it was the same.

So I thought that there is problem with nginx HTTP response. I compared
it completelly to lighttpd response, and made them exactly same. I even
changed the nginx server name to "lighttpd" (source recompilation was
needed), added headers like "X-Mod-H264-Streaming", but still no luck.

When serving video through lighttpd, it starts play immediately.
When serving through nginx, it starts play only after it is fully
loaded.

However requests for partial files like video.mp4?start=32.33 works well
with nginx. I think there is problem with initial request (the one
without "start" parameter).

I'm stuck. Any ideas what else should be checked?
Any advice would be appreciated. Thanks.

Ondrej

--
Ondrej JOMBIK
Platon Technologies s.r.o., Hlavna 3, Sala SK-92701
+421 903 PLATON - info at platon.org - http://platon.org




From reallfqq-nginx at yahoo.fr  Mon Jan 30 06:35:06 2012
From: reallfqq-nginx at yahoo.fr (B.R.)
Date: Mon, 30 Jan 2012 01:35:06 -0500
Subject: MP4 (H264) migrating from lighttpd to nginx
In-Reply-To: <alpine.DEB.2.00.1201300400040.29347@athena.platon.sk>
References: <alpine.DEB.2.00.1201300400040.29347@athena.platon.sk>
Message-ID: <CALqce=3JEOw-36xROzbgGf5rSyi=TLRx1GUgOFAdgUpWXLyB7g@mail.gmail.com>

I am no expert in streaming, but I read from the
ngx_http_mp4_module<http://nginx.org/en/docs/http/ngx_http_mp4_module.html>page
a special request of 'start=0' would transfer metadata.
Maybe without it, the file would simply be considered as a normal file and
then served completely before being able to be read ?

My 2 cents...
---
*B. R.*


On Sun, Jan 29, 2012 at 22:09, Ondrej Jombik <jombik at platon.org> wrote:

> It is great that MP4 module came out. So we decided to migrate our video
> streaming servers to nginx. In fact those are our last non-nginx
> webservers (not counting IIS).
>
> However we have some troubles. In our Flash player, MP4 video starts
> play only after it is completelly loaded. I thought there is problem
> with player, so I tried open-source flowplayer and it was the same.
>
> So I thought that there is problem with nginx HTTP response. I compared
> it completelly to lighttpd response, and made them exactly same. I even
> changed the nginx server name to "lighttpd" (source recompilation was
> needed), added headers like "X-Mod-H264-Streaming", but still no luck.
>
> When serving video through lighttpd, it starts play immediately.
> When serving through nginx, it starts play only after it is fully
> loaded.
>
> However requests for partial files like video.mp4?start=32.33 works well
> with nginx. I think there is problem with initial request (the one
> without "start" parameter).
>
> I'm stuck. Any ideas what else should be checked?
> Any advice would be appreciated. Thanks.
>
> Ondrej
>
> --
> Ondrej JOMBIK
> Platon Technologies s.r.o., Hlavna 3, Sala SK-92701
> +421 903 PLATON - info at platon.org - http://platon.org
>
>
>
> ______________________________**_________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/**mailman/listinfo/nginx<http://mailman.nginx.org/mailman/listinfo/nginx>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120130/25ea3b30/attachment.html>

From jombik at platon.org  Mon Jan 30 07:08:13 2012
From: jombik at platon.org (Ondrej Jombik)
Date: Mon, 30 Jan 2012 08:08:13 +0100 (CET)
Subject: MP4 (H264) migrating from lighttpd to nginx
In-Reply-To: <CALqce=3JEOw-36xROzbgGf5rSyi=TLRx1GUgOFAdgUpWXLyB7g@mail.gmail.com>
References: <alpine.DEB.2.00.1201300400040.29347@athena.platon.sk>
	<CALqce=3JEOw-36xROzbgGf5rSyi=TLRx1GUgOFAdgUpWXLyB7g@mail.gmail.com>
Message-ID: <alpine.DEB.2.00.1201300755340.29347@athena.platon.sk>

Thanks B.R., you are exactly right, after appending "?start=0" video
starts to play immediatelly.

I also discovered that our MP4 files are encoded with index at the end
of file. lighttpd was silently moving this index (moov) to the
beginning.

That raises question for some nginx performance guru:

What is better: recode all the MP4 files (thousands of thousands) and
move index to the beginning or simply use "?start=0" parameter?

I feel that recoding may save us some memory during future request
processing, because with "?start=0" nginx must at least see the end of
file and fetch the index.

Any opinion would be welcome. I have no problem to recode/reindex milion
files, but it has to have some meaning first ;-)

Thanks

Ondrej

On Mon, 30 Jan 2012, B.R. wrote:

> I am no expert in streaming, but I read from the
> ngx_http_mp4_module<http://nginx.org/en/docs/http/ngx_http_mp4_module.html>page
> a special request of 'start=0' would transfer metadata.
> Maybe without it, the file would simply be considered as a normal file and
> then served completely before being able to be read ?
>
> My 2 cents...
> ---
> *B. R.*
>
>
> On Sun, Jan 29, 2012 at 22:09, Ondrej Jombik <jombik at platon.org> wrote:
>
>> It is great that MP4 module came out. So we decided to migrate our video
>> streaming servers to nginx. In fact those are our last non-nginx
>> webservers (not counting IIS).
>>
>> However we have some troubles. In our Flash player, MP4 video starts
>> play only after it is completelly loaded. I thought there is problem
>> with player, so I tried open-source flowplayer and it was the same.
>>
>> So I thought that there is problem with nginx HTTP response. I compared
>> it completelly to lighttpd response, and made them exactly same. I even
>> changed the nginx server name to "lighttpd" (source recompilation was
>> needed), added headers like "X-Mod-H264-Streaming", but still no luck.
>>
>> When serving video through lighttpd, it starts play immediately.
>> When serving through nginx, it starts play only after it is fully
>> loaded.
>>
>> However requests for partial files like video.mp4?start=32.33 works well
>> with nginx. I think there is problem with initial request (the one
>> without "start" parameter).
>>
>> I'm stuck. Any ideas what else should be checked?
>> Any advice would be appreciated. Thanks.
>>
>> Ondrej
>>
>> --
>> Ondrej JOMBIK
>> Platon Technologies s.r.o., Hlavna 3, Sala SK-92701
>> +421 903 PLATON - info at platon.org - http://platon.org
>>
>>
>>
>> ______________________________**_________________
>> nginx mailing list
>> nginx at nginx.org
>> http://mailman.nginx.org/**mailman/listinfo/nginx<http://mailman.nginx.org/mailman/listinfo/nginx>
>>
>

--
Ondrej JOMBIK
Platon Technologies s.r.o., Hlavna 3, Sala SK-92701
+421 903 PLATON - info at platon.org - http://platon.org




From mdounin at mdounin.ru  Mon Jan 30 07:31:10 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 30 Jan 2012 11:31:10 +0400
Subject: MP4 (H264) migrating from lighttpd to nginx
In-Reply-To: <alpine.DEB.2.00.1201300400040.29347@athena.platon.sk>
References: <alpine.DEB.2.00.1201300400040.29347@athena.platon.sk>
Message-ID: <20120130073109.GP67687@mdounin.ru>

Hello!

On Mon, Jan 30, 2012 at 04:09:33AM +0100, Ondrej Jombik wrote:

> It is great that MP4 module came out. So we decided to migrate our video
> streaming servers to nginx. In fact those are our last non-nginx
> webservers (not counting IIS).
> 
> However we have some troubles. In our Flash player, MP4 video starts
> play only after it is completelly loaded. I thought there is problem
> with player, so I tried open-source flowplayer and it was the same.
> 
> So I thought that there is problem with nginx HTTP response. I compared
> it completelly to lighttpd response, and made them exactly same. I even
> changed the nginx server name to "lighttpd" (source recompilation was
> needed), added headers like "X-Mod-H264-Streaming", but still no luck.
> 
> When serving video through lighttpd, it starts play immediately.
> When serving through nginx, it starts play only after it is fully
> loaded.
> 
> However requests for partial files like video.mp4?start=32.33 works well
> with nginx. I think there is problem with initial request (the one
> without "start" parameter).
> 
> I'm stuck. Any ideas what else should be checked?
> Any advice would be appreciated. Thanks.

On initial request nginx returns original file, and if moov atom 
with metadata goes after mdat (one with media data) you'll see 
"video starts after loading" problem.  Recoding files to place 
moov atom first will solve it.

Maxim Dounin


From mike503 at gmail.com  Mon Jan 30 07:51:19 2012
From: mike503 at gmail.com (Michael Shadle)
Date: Sun, 29 Jan 2012 23:51:19 -0800
Subject: MP4 (H264) migrating from lighttpd to nginx
In-Reply-To: <20120130073109.GP67687@mdounin.ru>
References: <alpine.DEB.2.00.1201300400040.29347@athena.platon.sk>
	<20120130073109.GP67687@mdounin.ru>
Message-ID: <D0A36345-FA8E-46C9-9084-A7E480068BFA@gmail.com>

I suggest using "qt-faststart" which is a simple command line tool which does this.

On Jan 29, 2012, at 11:31 PM, Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
> 
> On Mon, Jan 30, 2012 at 04:09:33AM +0100, Ondrej Jombik wrote:
> 
>> It is great that MP4 module came out. So we decided to migrate our video
>> streaming servers to nginx. In fact those are our last non-nginx
>> webservers (not counting IIS).
>> 
>> However we have some troubles. In our Flash player, MP4 video starts
>> play only after it is completelly loaded. I thought there is problem
>> with player, so I tried open-source flowplayer and it was the same.
>> 
>> So I thought that there is problem with nginx HTTP response. I compared
>> it completelly to lighttpd response, and made them exactly same. I even
>> changed the nginx server name to "lighttpd" (source recompilation was
>> needed), added headers like "X-Mod-H264-Streaming", but still no luck.
>> 
>> When serving video through lighttpd, it starts play immediately.
>> When serving through nginx, it starts play only after it is fully
>> loaded.
>> 
>> However requests for partial files like video.mp4?start=32.33 works well
>> with nginx. I think there is problem with initial request (the one
>> without "start" parameter).
>> 
>> I'm stuck. Any ideas what else should be checked?
>> Any advice would be appreciated. Thanks.
> 
> On initial request nginx returns original file, and if moov atom 
> with metadata goes after mdat (one with media data) you'll see 
> "video starts after loading" problem.  Recoding files to place 
> moov atom first will solve it.
> 
> Maxim Dounin
> 
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From igor at sysoev.ru  Mon Jan 30 08:00:39 2012
From: igor at sysoev.ru (Igor Sysoev)
Date: Mon, 30 Jan 2012 12:00:39 +0400
Subject: MP4 (H264) migrating from lighttpd to nginx
In-Reply-To: <alpine.DEB.2.00.1201300755340.29347@athena.platon.sk>
References: <alpine.DEB.2.00.1201300400040.29347@athena.platon.sk>
	<CALqce=3JEOw-36xROzbgGf5rSyi=TLRx1GUgOFAdgUpWXLyB7g@mail.gmail.com>
	<alpine.DEB.2.00.1201300755340.29347@athena.platon.sk>
Message-ID: <20120130080039.GA85837@nginx.com>

On Mon, Jan 30, 2012 at 08:08:13AM +0100, Ondrej Jombik wrote:
> Thanks B.R., you are exactly right, after appending "?start=0" video
> starts to play immediatelly.
> 
> I also discovered that our MP4 files are encoded with index at the end
> of file. lighttpd was silently moving this index (moov) to the
> beginning.
> 
> That raises question for some nginx performance guru:
> 
> What is better: recode all the MP4 files (thousands of thousands) and
> move index to the beginning or simply use "?start=0" parameter?
> 
> I feel that recoding may save us some memory during future request
> processing, because with "?start=0" nginx must at least see the end of
> file and fetch the index.
> 
> Any opinion would be welcome. I have no problem to recode/reindex milion
> files, but it has to have some meaning first ;-)

>From perfomance point of view it's better to move the index to the beginning.
However, you do not need to recode the files, but just to move the index.


-- 
Igor Sysoev


From varia at e-healthexpert.org  Mon Jan 30 09:23:38 2012
From: varia at e-healthexpert.org (Mark Alan)
Date: Mon, 30 Jan 2012 09:23:38 +0000
Subject: MP4 (H264) migrating from lighttpd to nginx
In-Reply-To: <D0A36345-FA8E-46C9-9084-A7E480068BFA@gmail.com>
References: <alpine.DEB.2.00.1201300400040.29347@athena.platon.sk>
	<20120130073109.GP67687@mdounin.ru>
	<D0A36345-FA8E-46C9-9084-A7E480068BFA@gmail.com>
Message-ID: <20120130092338.4e722669@e-healthexpert.org>

On Sun, 29 Jan 2012 23:51:19 -0800, Michael Shadle <mike503 at gmail.com>
wrote:

> I suggest using "qt-faststart" which is a simple command line tool
> which does this.

A sort of qt-faststart howto:
http://www.stoimen.com/blog/2010/11/12/how-to-make-mp4-progressive-with-qt-faststart/


M.


From reallfqq-nginx at yahoo.fr  Mon Jan 30 11:34:12 2012
From: reallfqq-nginx at yahoo.fr (B.R.)
Date: Mon, 30 Jan 2012 06:34:12 -0500
Subject: MP4 (H264) migrating from lighttpd to nginx
In-Reply-To: <20120130092338.4e722669@e-healthexpert.org>
References: <alpine.DEB.2.00.1201300400040.29347@athena.platon.sk>
	<20120130073109.GP67687@mdounin.ru>
	<D0A36345-FA8E-46C9-9084-A7E480068BFA@gmail.com>
	<20120130092338.4e722669@e-healthexpert.org>
Message-ID: <CALqce=2yxMp14M6s8ntNcybYCsHUTXh60iG0wMf-jExR1mczbg@mail.gmail.com>

Hi again,

I am glad a little Googling + some basic reading made me able to answer
your expert question! :oP

I am still no expert, but from what I read on the link I gave you, there is
a link to how to prepare files to be served in streaming.
It was mentioned that if a file is not prepared as it should be (metadata
at the start of files), the consequence was the calculation of this missing
metadata CPU & I/O overhead, including extra harddisk use... which is bad
for performance! They even not mention the cache faults...
It is said that the metadata must be the first thing read in a file which
is read sequentially from the beginning. So, if your metadata is at the
end... I let you finish that ^^

Nginx, like lighttpd, is able to do that on its own but that kills
performance, so the better the files are prepared, the less you will use
resources on your server.
However, as Igor mentioned, it is only about moving metadata, not encoding
the whole files again, which is a simpler task to do.

If you wish to process your files and that you have a lot of them, maybe it
is not worth it to do so with a unique and long shot.
Maybe you could make your files served through some checking server-sided
script, which will serve the data from the current file and concurrently
regenerate a new well-made file for future requests?
That's maybe not the best solution but that would allow to start converting
regularly served files first. Of course it would be a temporary one...
---
*B. R.*


On Mon, Jan 30, 2012 at 04:23, Mark Alan <varia at e-healthexpert.org> wrote:

> On Sun, 29 Jan 2012 23:51:19 -0800, Michael Shadle <mike503 at gmail.com>
> wrote:
>
> > I suggest using "qt-faststart" which is a simple command line tool
> > which does this.
>
> A sort of qt-faststart howto:
>
> http://www.stoimen.com/blog/2010/11/12/how-to-make-mp4-progressive-with-qt-faststart/
>
>
> M.
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx/attachments/20120130/4db9ed49/attachment.html>

From mdounin at mdounin.ru  Mon Jan 30 14:54:21 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Mon, 30 Jan 2012 18:54:21 +0400
Subject: nginx-1.1.14
Message-ID: <20120130145421.GB67687@mdounin.ru>

Changes with nginx 1.1.14                                        30 Jan 2012

    *) Feature: multiple "limit_req" limits may be used simultaneously.

    *) Bugfix: in error handling while connecting to a backend.
       Thanks to Piotr Sikora.

    *) Bugfix: in AIO error handling on FreeBSD.

    *) Bugfix: in the OpenSSL library initialization.

    *) Bugfix: the "proxy_redirect" directives might not be correctly
       inherited.

    *) Bugfix: memory leak during reconfiguration if the "pcre_jit"
       directive was used.


Maxim Dounin


From jombik at platon.org  Mon Jan 30 16:09:19 2012
From: jombik at platon.org (Ondrej Jombik)
Date: Mon, 30 Jan 2012 17:09:19 +0100 (CET)
Subject: MP4 (H264) migrating from lighttpd to nginx
In-Reply-To: <CALqce=2yxMp14M6s8ntNcybYCsHUTXh60iG0wMf-jExR1mczbg@mail.gmail.com>
References: <alpine.DEB.2.00.1201300400040.29347@athena.platon.sk>
	<20120130073109.GP67687@mdounin.ru>
	<D0A36345-FA8E-46C9-9084-A7E480068BFA@gmail.com>
	<20120130092338.4e722669@e-healthexpert.org>
	<CALqce=2yxMp14M6s8ntNcybYCsHUTXh60iG0wMf-jExR1mczbg@mail.gmail.com>
Message-ID: <alpine.DEB.2.00.1201301656330.29347@athena.platon.sk>

Sometimes it happens that you are working under light, and that light is
so bright, so you are unable to see obvious things. This happened to me.

Yes, now my initial question looks a bit strange, but thanks to asking
I have at least two options how to deal with the issue.

According to Igor, moving index to the beginning of the file should be
better from the performance point of view. Since we are migrating not
only from lighttpd to nginx, but also to new storage, I will put this
"recoding" into migration batch and we will have nice MP4 files at the end.

Thanks again and... nginx mp4 module rocks! :)

Ondrej

On Mon, 30 Jan 2012, B.R. wrote:

> Hi again,
>
> I am glad a little Googling + some basic reading made me able to answer
> your expert question! :oP
>
> I am still no expert, but from what I read on the link I gave you, there is
> a link to how to prepare files to be served in streaming.
> It was mentioned that if a file is not prepared as it should be (metadata
> at the start of files), the consequence was the calculation of this missing
> metadata CPU & I/O overhead, including extra harddisk use... which is bad
> for performance! They even not mention the cache faults...
> It is said that the metadata must be the first thing read in a file which
> is read sequentially from the beginning. So, if your metadata is at the
> end... I let you finish that ^^
>
> Nginx, like lighttpd, is able to do that on its own but that kills
> performance, so the better the files are prepared, the less you will use
> resources on your server.
> However, as Igor mentioned, it is only about moving metadata, not encoding
> the whole files again, which is a simpler task to do.
>
> If you wish to process your files and that you have a lot of them, maybe it
> is not worth it to do so with a unique and long shot.
> Maybe you could make your files served through some checking server-sided
> script, which will serve the data from the current file and concurrently
> regenerate a new well-made file for future requests?
> That's maybe not the best solution but that would allow to start converting
> regularly served files first. Of course it would be a temporary one...
> ---
> *B. R.*
>
>
> On Mon, Jan 30, 2012 at 04:23, Mark Alan <varia at e-healthexpert.org> wrote:
>
>> On Sun, 29 Jan 2012 23:51:19 -0800, Michael Shadle <mike503 at gmail.com>
>> wrote:
>>
>>> I suggest using "qt-faststart" which is a simple command line tool
>>> which does this.
>>
>> A sort of qt-faststart howto:
>>
>> http://www.stoimen.com/blog/2010/11/12/how-to-make-mp4-progressive-with-qt-faststart/

--
Ondrej JOMBIK
Platon Technologies s.r.o., Hlavna 3, Sala SK-92701
+421 903 PLATON - info at platon.org - http://platon.org




From nginx-forum at nginx.us  Mon Jan 30 17:50:56 2012
From: nginx-forum at nginx.us (chrislovecnm)
Date: Mon, 30 Jan 2012 12:50:56 -0500
Subject: Module Advice - Cassandra / Thrift
Message-ID: <0c6891000f475719ea484a407ee60e1d.NginxMailingListEnglish@forum.nginx.org>

Hi all ... looking for some advice.

I am looking to use the lau module in conjunction with a module that
talks to nosql cassandra database. A couple of questions:

Does a module exist that can talk to Apache Cassandra NoSQL db?  I am
thinking no.

Cassandra uses Apache Thrift as it's driver / service layer.  Using
Thrift you can write in C, but then you have a glib dependency, or I can
code in C++.  What do you recommend?  Adding the glib dependency or
writing an upstream implementation that talks to a C++ library that is
compatible with C?

Thanks

Chris

P.S. I am planning on making this opensource ...

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221807,221807#msg-221807


From agentzh at gmail.com  Tue Jan 31 04:05:00 2012
From: agentzh at gmail.com (agentzh)
Date: Tue, 31 Jan 2012 12:05:00 +0800
Subject: Module Advice - Cassandra / Thrift
In-Reply-To: <0c6891000f475719ea484a407ee60e1d.NginxMailingListEnglish@forum.nginx.org>
References: <0c6891000f475719ea484a407ee60e1d.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <CAB4Tn6N1-3FjeJUQwVSS68gKppSn8Rxz2e0PFdZHExUbvtXCYQ@mail.gmail.com>

On Tue, Jan 31, 2012 at 1:50 AM, chrislovecnm <nginx-forum at nginx.us> wrote:
> Hi all ... looking for some advice.
>
> I am looking to use the lau module in conjunction with a module that
> talks to nosql cassandra database. A couple of questions:
>

I assume you're referring to the lua module (I have no idea what lau module is).

> Does a module exist that can talk to Apache Cassandra NoSQL db? ?I am
> thinking no.
>

The ngx_lua module's cosocket branch has implemented a nonblocking but
still synchronous TCP/unix domain socket API for Lua. I think the most
natural and efficient way is to code up a TCP client library in pure
Lua that understands Cassandra's wired protocol if I'm not missing
anything obvious here :)

> Cassandra uses Apache Thrift as it's driver / service layer. ?Using
> Thrift you can write in C, but then you have a glib dependency, or I can
> code in C++. ?What do you recommend? ?Adding the glib dependency or
> writing an upstream implementation that talks to a C++ library that is
> compatible with C?
>

Reusing existing Cassandra drivers could be much trickier compared to
cosocket-based pure Lua drivers as mentioned above, but Piotr Sikora
has done something quite generic in his ngx_zeromq module, maybe Piotr
can give some advice here.

Providing a hook for native 3rd-party drivers in ngx_lua's cosocket
mechanism is also on my TODO list though :)

>
> P.S. I am planning on making this opensource ...
>

Yay!

Best,
-agentzh


From nginx-forum at nginx.us  Tue Jan 31 04:41:35 2012
From: nginx-forum at nginx.us (chrislovecnm)
Date: Mon, 30 Jan 2012 23:41:35 -0500
Subject: Module Advice - Cassandra / Thrift
In-Reply-To: <CAB4Tn6N1-3FjeJUQwVSS68gKppSn8Rxz2e0PFdZHExUbvtXCYQ@mail.gmail.com>
References: <CAB4Tn6N1-3FjeJUQwVSS68gKppSn8Rxz2e0PFdZHExUbvtXCYQ@mail.gmail.com>
Message-ID: <4d16f8f75627cf62f881d1556158005c.NginxMailingListEnglish@forum.nginx.org>

Thanks Agentzh,

I initially wanted to right a direct socket connection with Cassandra,
but their protocol is not well documented.  Looks like I get to figure
out how Thrift works.  It may also may make upgrades a bit tricky :D

Bottom line is that I need high performance, and I am really concerned
that Thrift is going to get in the way.

Well let me take a look figuring out what the heck this Thrift client is
actually doing.

Thanks for the advice.  I am sure it is not going to be the last time I
ask.

Now the true question is lua or c ...

Chris

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221807,221819#msg-221819


From yoann at queret.net  Tue Jan 31 10:16:37 2012
From: yoann at queret.net (Yoann QUERET)
Date: Tue, 31 Jan 2012 11:16:37 +0100
Subject: HttpRewriteModule and use more than 10 variables
Message-ID: <4F27BF85.5090904@queret.net>

Hi,

In a rewrite, is it possible to use more than 9 variables ?

How i can use $10, $11, ... ?

rewrite 
"^/img/([0-9a-zA-Z]{1})([0-9a-zA-Z]{7})([0-9a-zA-Z]{1})([0-9a-zA-Z]{7})/([0-9]{4})([0-9]{2})([0-9]{2})/([0-9a-fA-F]{8})([0-9a-fA-F]{4})([0-9a-fA-F]{4})([0-9a-fA-F]{4})([0-9a-fA-F]{12}).*$" 
/$1/$2/$3/$4/$5/$6/$7/$8/$9/$10/$11/$12.image;

Thk


From kworthington at gmail.com  Tue Jan 31 10:41:25 2012
From: kworthington at gmail.com (Kevin Worthington)
Date: Tue, 31 Jan 2012 05:41:25 -0500
Subject: nginx-1.1.14
In-Reply-To: <20120130145421.GB67687@mdounin.ru>
References: <20120130145421.GB67687@mdounin.ru>
Message-ID: <CAGo79UV1ULA=GqK2nZBMYm-9sdvOy=CE53p2UnfXuZgkU+wXag@mail.gmail.com>

Hello Nginx Users,

Now available: Nginx 1.1.14 For Windows http://goo.gl/hWymA (32-bit
and 64-bit versions)

These versions are to support legacy users who are already using
Cygwin based builds of Nginx. Officially supported native Windows
binaries are at nginx.org.

Thank you,
Kevin
--
Kevin Worthington
kworthington *@~ #gmail}  [dot) {com]
http://www.kevinworthington.com/


On Mon, Jan 30, 2012 at 9:54 AM, Maxim Dounin <mdounin at mdounin.ru> wrote:
> Changes with nginx 1.1.14 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?30 Jan 2012
>
> ? ?*) Feature: multiple "limit_req" limits may be used simultaneously.
>
> ? ?*) Bugfix: in error handling while connecting to a backend.
> ? ? ? Thanks to Piotr Sikora.
>
> ? ?*) Bugfix: in AIO error handling on FreeBSD.
>
> ? ?*) Bugfix: in the OpenSSL library initialization.
>
> ? ?*) Bugfix: the "proxy_redirect" directives might not be correctly
> ? ? ? inherited.
>
> ? ?*) Bugfix: memory leak during reconfiguration if the "pcre_jit"
> ? ? ? directive was used.
>
>
> Maxim Dounin
>
> _______________________________________________
> nginx mailing list
> nginx at nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx


From soulfury at simoncpu.com  Tue Jan 31 11:41:04 2012
From: soulfury at simoncpu.com (Simon Cornelius P. Umacob)
Date: Tue, 31 Jan 2012 19:41:04 +0800
Subject: IPv6 support for reverse proxy
Message-ID: <CAF=FVTtYCnyJ_7PXUr1UR4ZG-=_xkgZY_xXd+00ucukjJ69Dwg@mail.gmail.com>

Hi,

Can we use nginx to act as a reverse proxy for IPv6 servers?

i.e.,

server {
        listen [::]:80;
        listen 80;
        server_name  kame-experiment.net *.kame-experiment.net;

        location / {
                proxy_pass
http://[2001:200:dff:fff1:216:3eff:feb1:44d7];
                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;

        }
}


Thanks,

[ simon.cpu ]


From mp3geek at gmail.com  Tue Jan 31 11:45:43 2012
From: mp3geek at gmail.com (Ryan Brown)
Date: Wed, 1 Feb 2012 00:45:43 +1300
Subject: A low keepalive and limit_conn
Message-ID: <CAG_yTXzRFacJdOzNkf+MZ_XYqic4_snkS9i+EtSfY1a64Ncc7Q@mail.gmail.com>

Using keepalive (on zero) and connection limit, will the limit_conn
ever be hit if I had a low keepalive?

keepalive_timeout  0;
limit_conn_zone   $binary_remote_addr  zone=addr:10m;
limit_conn   addr   10;


From ne at vbart.ru  Tue Jan 31 12:38:19 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Tue, 31 Jan 2012 16:38:19 +0400
Subject: HttpRewriteModule and use more than 10 variables
In-Reply-To: <4F27BF85.5090904@queret.net>
References: <4F27BF85.5090904@queret.net>
Message-ID: <201201311638.19973.ne@vbart.ru>

On Tuesday 31 January 2012 14:16:37 Yoann QUERET wrote:
> Hi,
> 
> In a rewrite, is it possible to use more than 9 variables ?

Yes, it's possible.

> How i can use $10, $11, ... ?
> 

Use named captures instead.

 wbr, Valentin V. Bartenev


From nginx-forum at nginx.us  Tue Jan 31 12:46:46 2012
From: nginx-forum at nginx.us (pamribeirox)
Date: Tue, 31 Jan 2012 07:46:46 -0500
Subject: ip_hash load balancer incompatible with IPv6 compat. sockets
Message-ID: <62c4ac6e5141667824b3c6def0184571.NginxMailingListEnglish@forum.nginx.org>

I'm using nginx to distribute the load over 3 webservers using the
ip_hash method.
When the listening socket is IPv4/IPv6 compatible ( listen  [::]:443; )
only one of the backend servers are used.
With the socket in IPv4 only mode ( listen  0.0.0.0:443; ) the module
correctly distributes the load.
It seems the hash isn't taking in account the size of the addresses and
because the first 32bits of the struct in this mode doesn't seem to vary
much the load balacing isn't working.

Keep up with the good work with this excelent proxy!
Regards.

Posted at Nginx Forum: http://forum.nginx.org/read.php?2,221827,221827#msg-221827


From ne at vbart.ru  Tue Jan 31 12:49:25 2012
From: ne at vbart.ru (Valentin V. Bartenev)
Date: Tue, 31 Jan 2012 16:49:25 +0400
Subject: A low keepalive and limit_conn
In-Reply-To: <CAG_yTXzRFacJdOzNkf+MZ_XYqic4_snkS9i+EtSfY1a64Ncc7Q@mail.gmail.com>
References: <CAG_yTXzRFacJdOzNkf+MZ_XYqic4_snkS9i+EtSfY1a64Ncc7Q@mail.gmail.com>
Message-ID: <201201311649.25519.ne@vbart.ru>

On Tuesday 31 January 2012 15:45:43 Ryan Brown wrote:
> Using keepalive (on zero) and connection limit, will the limit_conn
> ever be hit if I had a low keepalive?

If client opens connection but doesn't send request, then limit_conn doesn't hit.

 wbr, Valentin V. Bartenev


From lists at ruby-forum.com  Tue Jan 31 13:24:50 2012
From: lists at ruby-forum.com (eric elg)
Date: Tue, 31 Jan 2012 14:24:50 +0100
Subject: rtsp over http response blocked by nginx
Message-ID: <2acf6c791d17ab41322379746c1e452e@ruby-forum.com>

Hello

I am trying to use NGINX as reverse proxy for streaming using QuickTime
and rtsp over http.
QuicKTime send an http GET to my reverse proxy in a public zone, then my
reverse proxy send the request to my video server in a private zone.
What I can see is that the http GET request is received by NGINX on the
reverse proxy, then by my video server, which responds immediately with
a 200 OK.
This response is received by the reverse proxy, but it seems that NGINX
block the response. (see log file attached).

Other http traffic is executed without any problem, problems occurs only
with rtsp over http.


The http headers of the http GET send by QuickTime are:
User-Agent: QuickTime/7.7 (qtver=7.7;os=Windows NT 5.1Service Pack 3)
x-sessioncookie: 4qqv28gOAABf4EQBBoAAAA
Accept: application/x-rtsp-tunnelled
Pragma: no-cache
Cache-Control: no-cache

Here is my NGINX configuration:

------- /etc/nginx/nginx.conf BEGIN -------
worker_processes  1;
error_log       /var/log/nginx/error.log ;

pid    /var/run/nginx/nginx.pid;

events {
    worker_connections  1024;
}

http {
    include       mime.types;
    default_type  application/octet-stream;
    access_log  /var/log/nginx/access.log;
    sendfile        on;

    keepalive_timeout  65;

    gzip            on;
    #gzip            off;
    gzip_min_length  1000;
    #gzip_buffers     4 8k;
    #gzip_http_version 1.0;
    gzip_comp_level 3;
    gzip_proxied any;
    #gzip_types      text/plain application/xml;
    gzip_types  text/plain text/css application/json
application/x-javascript text/xml application/xml application/xml+rss
text/javascript;
    gzip_vary         on;


   include /etc/nginx/sites-enabled/*;
   include /etc/nginx/proxy.conf;
}
------- /etc/nginx/nginx.conf END  -------


------- /etc/nginx/proxy.conf BEGIN  -------
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    10m;
client_body_buffer_size 128k;
proxy_connect_timeout   90;
proxy_send_timeout      90;
proxy_read_timeout      90;
proxy_buffers           32 4k;
------- /etc/nginx/proxy.conf END  -------


------- /etc/nginx/sites-enabled/video.conf BEGIN  -------
upstream video
{
    server 10.194.188.2:80;
}
server
{
    listen 10.194.188.90:80;
    server_name <public-ip>;
    location /
        {
         proxy_pass http://video;
   }
        error_log /var/log/nginx/errori debug;
}
------- /etc/nginx/sites-enabled/video.conf END  -------



Thank's for your help...

Eric

Attachments:
http://www.ruby-forum.com/attachment/6942/nginx.log


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


From mdounin at mdounin.ru  Tue Jan 31 14:28:44 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 31 Jan 2012 18:28:44 +0400
Subject: IPv6 support for reverse proxy
In-Reply-To: <CAF=FVTtYCnyJ_7PXUr1UR4ZG-=_xkgZY_xXd+00ucukjJ69Dwg@mail.gmail.com>
References: <CAF=FVTtYCnyJ_7PXUr1UR4ZG-=_xkgZY_xXd+00ucukjJ69Dwg@mail.gmail.com>
Message-ID: <20120131142844.GK67687@mdounin.ru>

Hello!

On Tue, Jan 31, 2012 at 07:41:04PM +0800, Simon Cornelius P. Umacob wrote:

> Hi,
> 
> Can we use nginx to act as a reverse proxy for IPv6 servers?
> 
> i.e.,
> 
> server {
>         listen [::]:80;
>         listen 80;
>         server_name  kame-experiment.net *.kame-experiment.net;
> 
>         location / {
>                 proxy_pass http://[2001:200:dff:fff1:216:3eff:feb1:44d7];

No, it's not currently supported.

Maxim Dounin


From mdounin at mdounin.ru  Tue Jan 31 15:19:30 2012
From: mdounin at mdounin.ru (Maxim Dounin)
Date: Tue, 31 Jan 2012 19:19:30 +0400
Subject: ip_hash load balancer incompatible with IPv6 compat. sockets
In-Reply-To: <62c4ac6e5141667824b3c6def0184571.NginxMailingListEnglish@forum.nginx.org>
References: <62c4ac6e5141667824b3c6def0184571.NginxMailingListEnglish@forum.nginx.org>
Message-ID: <20120131151930.GL67687@mdounin.ru>

Hello!

On Tue, Jan 31, 2012 at 07:46:46AM -0500, pamribeirox wrote:

> I'm using nginx to distribute the load over 3 webservers using the
> ip_hash method.
> When the listening socket is IPv4/IPv6 compatible ( listen  [::]:443; )
> only one of the backend servers are used.
> With the socket in IPv4 only mode ( listen  0.0.0.0:443; ) the module
> correctly distributes the load.
> It seems the hash isn't taking in account the size of the addresses and
> because the first 32bits of the struct in this mode doesn't seem to vary
> much the load balacing isn't working.

Yes, ip_hash only works with ipv4 addresses, it's not currently 
able to handle ipv6 addresses (including v4-mapped ones).

Maxim Dounin