Apache overloads server

R

ruroshin

New member
Joined
Mar 24, 2005
Messages
2
Hi,

I'm running DA on Fedora Core 3, Apache 1.3.33.

For the past week or so Apache would sporadically go nuts and use up all servers resources to the point that load goes past 100+. At first I thought it might of been a DoS attack but after a reboot its fine again, if it was a DoS attack I would assume the load would shoot up again after the reboot.

Looking at the Apache error log I see this:

[Thu Mar 24 16:00:32 2005] [error] could not make child process 19855 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19857 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19859 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 20447 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19861 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19862 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 20417 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19865 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19866 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19867 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 20464 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19869 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 20472 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19872 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19873 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 20473 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19875 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19876 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19878 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [error] could not make child process 19879 exit, attempting to continue anyway
[Thu Mar 24 16:00:32 2005] [notice] caught SIGTERM, shutting down
[Thu Mar 24 16:00:44 2005] [notice] Apache/1.3.33 (Unix) mod_ssl/2.8.22 OpenSSL/0.9.7a PHP/4.3.10 mod_perl/1.29 FrontPage/5.0.2.2510 configured -- resuming normal operations
[Thu Mar 24 16:00:44 2005] [notice] suEXEC mechanism enabled (wrapper: /usr/sbin/suexec)
[Thu Mar 24 16:00:44 2005] [notice] Accept mutex: sysvsem (Default: sysvsem)
[Thu Mar 24 16:16:00 2005] [notice] child pid 22471 exit signal Bus error (7)
[Thu Mar 24 16:16:52 2005] [error] server reached MaxClients setting, consider raising the MaxClients setting
*** glibc detected *** double free or corruption (!prev): 0x092b7668 ***

I checked my maxclients setting and its at 450.

Any ideas on whats going on?
 
Last edited:
I had this issue a few days back too. Load went up to about 60 (!), even SSH access was almost impossible.

Within a few minutes it was over though...
 
I have it too

Every day during last week at 5 o’clock appache overloads server then crashes mysql I cant understand whats wrong.

cronjob is clean there is only DA jobs.

i'he FreeBSD 5.3 appache 1.33 mysql 4.017 (i have not changed what the DA installed)
I have only upgrade the DA from 1.23.X to 1.24.2

PLEASE HELP
 
You might want to try upgrading MySQL to a more recent version first and then see if the problem still persists. The fact that it's happening everyday at a specific time could lead the realization of a cronjob doing it. What cronjobs run at or around 5 (including directadmin's?) for example
 
Xandrios said:
I had this issue a few days back too. Load went up to about 60 (!), even SSH access was almost impossible.

Within a few minutes it was over though...
Click to expand...
how do you know what your load is? where do you get 60?
 
I am getting the same thing heres my logs

Apache/1.3.33
Zend Engine v1.3.0
PHP 4.3.11
mysql Ver 12.22 Distrib 4.0.23,

http error log

[Mon Jun 6 20:50:12 2005] [error] could not make child process 15615 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 11587 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 15618 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 15619 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 11597 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14299 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14351 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14357 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14358 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 10557 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 12745 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 12746 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14359 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 15733 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 12858 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 15841 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14364 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14461 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 12868 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 11618 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 15842 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 15843 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 15844 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 15846 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 11693 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 15847 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14471 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14608 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 15848 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14474 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14611 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14615 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 12988 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14617 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14618 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 12997 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 12998 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14622 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 15850 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14624 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14626 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14683 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14753 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14757 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 11751 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [error] could not make child process 14870 exit, attempting to continue anyway
[Mon Jun 6 20:50:12 2005] [notice] caught SIGTERM, shutting down
[Mon Jun 6 20:53:40 2005] [notice] Apache/1.3.33 (Unix) mod_ssl/2.8.22 OpenSSL/0.9.7a PHP/4.3.11 mod_perl/1.29 FrontPage/5.0.2.2510 configured -- resuming normal operations
[Mon Jun 6 20:53:40 2005] [notice] suEXEC mechanism enabled (wrapper: /usr/sbin/suexec)
[Mon Jun 6 20:53:40 2005] [notice] Accept mutex: sysvsem (Default: sysvsem)
Click to expand...

any idea whats causing this?
 
Generally the problem is a DOS attack.

We've had to shut down httpd and restart it a few times in the last month; normally we don't have to do it at all.

Jeff
 
dos attac

Maybe it --s a dos attac but i can't monitor all servers,
to restart apache, whole day. its too hard!
 
If you don't monitor your servers you'll probably lose business to those of us who do.

Your options are to not monitor your servers, to monitor them manually, to use monitoring software (much of it free) you can install (such as nagios, for example), have a company monitor for you, or have your clients monitor for you by notifying you when your servers are down.

Jeff
 
You must log in or register to reply here.
Back
Top