Debugging mysterious VDS reboots in CentOS

[moiph]

I've got a VDS running CentOS 5.6, setup as a single core web server running cpanel/apache/mysql etc. In the past week it's rebooted 3 times -- extremely unexpected and undesirable.

Any tips on where to look through system logs to figure out what's going on?

Looking through apache logs I see various 404s and then

Code: Select all

[Thu Aug 11 23:19:33 2011] [warn] pid file /usr/local/apache/logs/httpd.pid overwritten -- Unclean shutdown of previous Apache run?

So there was definitely an unexpected shutdown. But so far I have no idea what...

[Edge100x]

/var/log/messages would be the main place to check. Ideally, there will be kernel output there saying what happened at the time of the crash.

Are you running the latest stable kernel release?

[moiph]

I'm not running the latest kernel no, updating now.

In /var/log/messages I do see:

Aug 11 23:18:04 kelvin syslogd 1.4.1: restart.

Indicating a restart...but no clues as to why. I do have cpanel set to handle upgrading software but I'm not aware of it being able to kick off system reboots...

[moiph]

I don't see an edit so I guess I'll double post

Updated the kernel, and did a manual reboot...and now comparing the logs my reboot has much more verbose logs with various services shutting down, etc. Whereas the unexpected reboot was just the sudden restart. So I guess that would rule out my shot in the dark about cpanel kicking off a reboot