Apache keep catching SIGTERM out of the blue

Asked Dec 05 '14 at 12:57

Active Dec 05 '14 at 12:57

Viewed 158 times

I have apache 2.2.15 server on CentOS release 6.6 (Final). It runs mod_perl and some perl scripts in junction with mysql DB. The problem I'm encountering is that from time to time apache gets SIGTERM (and it is written into its error log).

[notice] caught SIGTERM, shutting down

After that Apache is in half alive state. The processes present in the system. Usually httpd.pid is not present anymore. Visiting the site yields 403 Forbidden page on all URLs.

It happens both when the web site is in use or when on idle...

So what could be the reason why it is happening? How do I know whats happening inside there?

edited Dec 05 '14 at 12:57

MadHatter

78,442
20
178
229

asked Dec 05 '14 at 12:57

Unicorn

Look through your error logs. It could have been OOM killed or something. – Peter Dec 05 '14 at 13:55
@Peter, I checked the memory usage while server was alive using /proc/meminfo and kept getting same 300MBs of free RAM. The command output showed 0 SwapTotal and 0 SwapFree but there were plenty of RAM. – Unicorn Dec 05 '14 at 15:24
Output of `dmesg`? – Nathan C Dec 05 '14 at 16:08
@NathanC, I don't know `sudo dmesg` outputs nothing... – Unicorn Dec 05 '14 at 16:26
If you have strace installed, you could try to "catch it in the act." Something like: `strace -o myfilename -ffq -p ` – JasonAzze Dec 05 '14 at 21:21

Apache keep catching SIGTERM out of the blue

0 Answers0