How to catch error 500 on server with centos?

Question

MySQL

Good afternoon!
Help me understand the algorithm of actions when a 500 error occurs on the server.
The error appears every 20-30 minutes, the server goes down completely, neither ftp nor ssh is available, static files are also not given (robots.txt).
The control panel is cpanel(WHM).
In apache logs:

[Tue Jan 27 09:14:49.589345 2015] [mpm_prefork:notice] [pid 21856] AH00171: Graceful restart requested, doing restart
[Tue Jan 27 09:14:49.998472 2015] [ssl:warn] [pid 21856] AH01906: WHM.:443:0 server certificate is a CA certificate (BasicConstraints: CA == TRUE !?)
[Tue Jan 27 09:14:49.998537 2015] [ssl:warn] [pid 21856] AH01909: WHM.:443:0 server certificate does NOT include an ID which matches the server name
[Tue Jan 27 09:14:50.000548 2015] [mpm_prefork:notice] [pid 21856] AH00163: Apache/2.4.10 (Unix) OpenSSL/0.9.8e-fips-rhel5 mod_bwlimited/1.4 configured -- resuming normal operations
[Tue Jan 27 09:14:50.000588 2015] [core:notice] [pid 21856] AH00094: Command line: '/usr/local/apache/bin/httpd -D SSL'

The SSL certificate was deleted, the apache config was recreated, but errors still pour in. How to fix it?
upd. On the fifth day of jambs, the hoster admitted that there were problems with the hardware and the site would be replaced.
Thanks for answers!

Answer 1

Even if ssh goes down, then these are system problems, you need to look at /var/log/dmesg