Service down emails every few minutes on a fresh new server with no load (CentOS 7, kimsufi)

Ishware

Well-Known Member
Nov 7, 2003
211
6
168
Williamsburg, VA
cPanel Access Level
Root Administrator
Hi! :)

Got a new server via kimsufi, loaded the cPanel + CentOS 7 64-bit template.

All my services - imap, httpd, php_fpm - are failing nearly constantly. Getting emails approximately every 10 minutes - not exactly, it ranges from about 5-15 minutes. EDIT: The was true initially, but it changed to just apache (httpd) failing.

There's no load on the server - it's at zero right now, and has gone as high as about 1.5.

There's one account, a very minimal site (using apache header/footer to list a few files basically).

Httpd is failing constantly and not restarting. I can manually restart it via ssh (service httpd restart), and then it crashes in a few minutes or so (not immediately).

I'm not sure where to start with this. :/
 
Last edited:

Ishware

Well-Known Member
Nov 7, 2003
211
6
168
Williamsburg, VA
cPanel Access Level
Root Administrator
I have another server that started exhibiting the same behaviour, which is freaky as hell. On that server, I have a server management team in case of problems like this where I can't find the solution on my own and services are down.

Increasing the apache servers limit and max request workers by a factor of 10 seems to have resolved the issue for now.

For anyone finding this thread: WHM → Apache Configuration → General Configuration. Defaults are 256 and 150, respectively. Ended up increating to 2000 and 1500 respectively.

Anyone seeing this thread: Any idea why my other server with the same problem that cropped up tonight might've run fine for a number of months with moderately low load, then all of a sudden today with minimal load (offloaded some sites a couple of months ago) have this happen? I'd love some insight if there's an answer.
 

cPanelLauren

Product Owner
Staff member
Nov 14, 2017
13,295
1,272
313
Houston
What's noted in the apache error log for the new server? For the existing one any time you reach your limits with MaxRequestWorkers it will cause issue with Apache, this is clearly noted in the Apache error logs as well normally when it occurs.