Pingdom Uptime Problem

Portal Home > Knowledgebase > Articles Database > Pingdom Uptime Problem

Posted by kshazad86, 03-05-2012, 04:41 AM
My CentOS/WHM server recently went down according to Pingdom, but the problem is that I cant find out exactly what caused the 30 minutes of outage, this is what I have so far: - memory and cpu usage fine at the time - /var/log/messages isnt showing anything unusual other than FTP access occuring from 127.0.0.1 at the time of the outage - number of apache connections seems fine Any ideas what else I can check to find the cause of the downtime?
Posted by kshazad86, 03-05-2012, 05:04 AM
Also I got this error message from WHM: I've looked at the log, and this is what I get:
Posted by plumsauce, 03-05-2012, 07:07 AM
From your second post, your server really did die, so you do not have a "pingdom uptime problem". It really did happen.
Posted by kshazad86, 03-05-2012, 07:08 AM
Yes, exactly, but how can I find out the cause and how to make sure it doesnt happen again?
Posted by Patrick, 03-05-2012, 08:03 AM
So, did you raise MaxClients as the Apache error log suggested?
Posted by kshazad86, 03-05-2012, 08:04 AM
Yes, Maxclients has been raised from 150 to 200 now, although I want a permanent solution. Is there anyway of finding out if a certain user on the server generated a lot of proceses and caused the server to reach the limit?
Posted by Hoopla-Brad, 03-06-2012, 03:07 AM
The best solution would be limiting the problem before it happens, something like CloudLinux would help.
Posted by kshazad86, 03-06-2012, 04:45 AM
Funnily enough I'm already using CloudLinux on my server.
Posted by kshazad86, 03-06-2012, 05:36 AM
One other thing, if the server is flooded with "/usr/local/apache/bin/httpd -k start -DSSL" - then does this mean that alot of users/requests are coming in for a website via SSL? If so, is there anyway of seeing which IP's are using the SSL protocol on the server?