httpd-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Markus Mayer <>
Subject [users@httpd] finding the cause of a hung apache with 500+ child processes
Date Mon, 08 Sep 2008 11:00:05 GMT

I have an apache 2.0.61 with PHP 5.2.6 and the sunosin patches on Solaris 10, 
Sun V440 with 4 x 1.6GHz CPU's, 16Gb ram and 16Gb swap.  once or twice a year 
it hangs with hundreds of child processes.  The serverlimit is set to 512, 
and all the child processes are busy with some request that I can't figure 
out because I can't get a server-status from the server.  There is nothing 
written in the logs that I can find that would somehow indicate what has 
happened.  Suspucion was for a broken proxy, however there was no problem 
there.  Memory is also not a problem as no swap space (disk swap) was used at 
the time.  CPU load was in total less than 20%, mysql connections were also 
ok, at the time only 6.  There were also no nasty entries in the error logs.  
This apache server is otherwise very stable and responsive, usually running 
continuously until such a problem shows up.  The only way to get rid of this 
problem is to restart apache, then everything is ok, and there is no sign of 
problems anymore.  After restarting apache, I searched the log files for 
requests to the same url which looked like they might have taken a long time, 
but came up empty, it seems apache didn't log the open connections that it 
closed off at the restart.

Does anyone have any suggestions as to how I can pinpoint this problem?  Is 
there any way I can log requests before they are fully served by apache, as 
this might help the next time this shows up.

I would be grateful for any suggestions, and if anyone needs more information, 
please tell me.


The official User-To-User support forum of the Apache HTTP Server Project.
See <URL:> for more info.
To unsubscribe, e-mail:
   "   from the digest:
For additional commands, e-mail:

View raw message