Pantek Library
Hosting Provided By
CybrHost
High Speed Hosting

Apache2 crashes constantly on Etch

From: Aleks Feltin <aleksf(at)scms.fi>
Date: Mon Oct 08 2007 - 14:39:30 EDT


Hello,

Since the moment we have got an increased amount of customers, our server started to die once per a day (cronjob execution). Also, if using apache benchmark tool with a moderate amount of request and concurrent threads, apache2 processes die.

Environment I run, is following:

Debian GNU/Linux 4.0r1 (etch)

Apache/2.2.3 + PHP 5.2.0-8+etch7 (cgi-fcgi) + mod_fcgid.2.2

I didn't set any limits in limits.conf for user www-data

During the intensive execution apache fails with message in error.log:

sh: fork: Resource temporarily unavailable [Mon Oct 08 18:52:17 2007] [alert] (11)Resource temporarily unavailable: apr_thread_create: unable to create worker thread

Do you need help?X

After that I see no apache workers, and only php5-cgi running with uid=0 (normally run under uid=33). I cannot find explanation for such behavior, which drives me crazy, when server is down and some cgi processes are up with root privilleges :-/

My worker conf looks like following:

<IfModule mpm_worker_module>

    StartServers          2
    MaxClients          150
    MinSpareThreads      25
    MaxSpareThreads      75
    ThreadsPerChild      25

    MaxRequestsPerChild 0
</IfModule>

I did some experiments, changing it to:

<IfModule mpm_worker_module>

    ServerLimit          20
    StartServers          5
    MaxClients          500
    MinSpareThreads      25
    MaxSpareThreads      75
    ThreadsPerChild      25

    MaxRequestsPerChild 0
</IfModule>

Surprisingly, this change lead to error in error log.

[Sat Oct 06 14:35:21 2007] [alert] (11)Resource temporarily unavailable: apr_thread_create: unable to create worker thread

After direct php script execution (php5 $script) I see this message, when there are many execution threads:

Do you need more help?X

php (pre-forking): Resource temporarily unavailable

Seems some kind of limits prevent intensive forking and threading..I cannot figure that out. All my attepts has been failed to determine the cause. The same situation can be seen in our development server (load is lower compared to production), which is also Debian Stable.

If you have any thoughts or experience, please share it to me. I appreciate whatever kind of recommendations from you, guys. Hope for your assistance.

Thanks in advance!

-- 
Aleks Feltin
System Administrator
Sunset Wireless Ltd.

-- To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Received on Mon Oct 8 15:07:19 2007

This archive was generated by hypermail 2.1.8 : Mon Feb 25 2008 - 09:43:35 EST


Contact Us  Legal Notices  Order Services Online 
Pantek Home  Privacy Policy  IT news  Site Map  Pantek Library