no response nginx

shin fukuda albert at nn.iij4u.or.jp
Fri Apr 6 01:47:03 UTC 2012


Hello, Maxim.

On Thu, 5 Apr 2012 18:56:11 +0400
Maxim Dounin <mdounin at mdounin.ru> wrote:

> Hello!
> 
> On Thu, Apr 05, 2012 at 09:25:44AM +0900, shin fukuda wrote:
> 
> > Hi.
> > 
> > I have a problem with nginx-1.1.16 uses proxy_cache_lock on. 
> > When high load or many request to nginx, Suddenly nginx seems hang, 
> > no response at all. 
> > In netstat "SYN_RECV","CLOSE_WAIT" too many. This problem needs 
> > restarting nginx, and "CLOSE_WAIT" clear.
> > 
> > There is no problem in front-end nginx servers, but backends one.
> > 
> > Any clue?
> 
> I would suggest you are hitting some system limits.  Check nginx 
> and system logs, as well as various system stats.  Check the state 
> of nginx worker processes, most importantly wait channel (wchan, 
> "ps -alx" should show one at least on FreeBSD and Linux).

Set kernel parameter like 

 ...
 net.ipv4.tcp_syncookies = 1
 net.ipv4.tcp_tw_reuse = 1
 net.ipv4.tcp_max_syn_backlog = 8192
 net.ipv4.tcp_syncookies = 1
 net.ipv4.tcp_tw_recycle = 1
 net.ipv4.tcp_fin_timeout = 15
 net.ipv4.tcp_keepalive_time = 300

But, no good result so far.

Anyway, I check workers details, and find this,

 15832 write(4, "2012/04/04 19:56:10 [info] 15832#0: sigtimedwait() 
 failed (4: Interrupted system"..., 87) = 87
 15832 rt_sigtimedwait([ALRM IO RT_12 RT_13], 0x691220) = -1 EAGAIN 
 (Resource temporarily unavailable)
 15832 rt_sigtimedwait([ALRM IO RT_12 RT_13], 0x691220) = -1 EAGAIN 
 (Resource temporarily unavailable)
  ....
  ....

Is this normal status ?


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


-- 
shin fukuda <albert at nn.iij4u.or.jp>



More information about the nginx mailing list