Issues with 503 error and Plesk domains

[BrianDns]

I have been having a reoccurring issue over the past few days with certain domains timing out on my server. I have around 60 or 70 domains and subdomains total on the Plesk server running CentOS 6.2 and Plesk 10.4.4. Usually I was noticing the issues on random subdomains only within one domain name. But I started seeing the issue pop up more frequently on regular domains now too. Usually when it happens it is only on anywhere from 1 - 10 domains only. Not all of them. It is also seemingly random. When it happens I restart apache and everything goes back to normal for a half hour up to a few hours then it all starts up again.

I didnt see anything in the logs that looks out of the ordinary. There is a bunch of stuff about SSL errors for the Plesk Panel Cert. I recently removed the SSL cert from the panel.

Does anybody have any ideas what could be happening. Please Help!!!!!!

 

[abdi]

You sound to any of the following problems on your server:

1. Your apache runs out of child processes,

http://emergent.urbanpug.com/?p=60
https://www.virtacoresupport.com/index.php?_m=knowledgebase&_a=viewarticle&kbarticleid=102

2. Your server runs out of RAM (memory) and or SWAP.

Do you mind sharing here your server specs?

 

[BrianDns]

Issue resolved

Thank you for the response abdi. You were definitely on the right track. I resolved the issue late last night.

The culprit was a fastCGI setting.

The setting was: fastcgimaxprocesses. It was set to 20. Everytime the 21st website would try to start up a process it would get denied by the server. Simple fix, but just brain racking when your in the situation.

Thank you again for getting back quick!