This morning I woke up to one of my servers not responding to SSL, immediately my thoughts went to expired certificates – I checked and found I had 2 months left on contract, I took the time to renew, just in case. I then went back to the WHM Manager and reconfigured Apache – even so far as reinstalling but still the message “httpd failed to start please contact your sysadmin” continued.

As most Linux Geeks, I attempted to read logs – but nothing to be found. There’s nothing worse than not having a problem physically eminent on logs. After hours of basic linux commands through various Apache and cPanel logs: grep, tail etc… I was unable to locate the root of this issue. I manually started Apache from the command line and all went well – except there was still no SSL. ARGH!!

/usr/sbin/httpd configtest

Showed no errors, everything loaded fine. I went back to the WHM and reinstalled RPM packages – sat out the 20 minutes and still no SSL. httpd continued to fail on restart.

From here, it was evident that my httpd.conf was configured correctly – proof: Web Pages Loaded. SSL was the problem, and therefore I decided to shut down the ssl portion of Apache and try to debug from there. First I had to find the PID:

ps aux | grep httpd

What you are looking for here, is the PID for the SSL Apache Server. It’s easy to find this among all the other httpd lines. Simply look for the Ss.

kill -9 PID

I restarted this portion of the Apache Server

/usr/sbin/httpd sslstart

File Size Exceeds Limit…

Ah huh! Let’s find that offending file…

du -h /usr/local/apache/domlogs/*

This returned multiple files, at which 2 were way over 4 G. I issued the following commands for both files.

rm domainname.ssl_data_log

Restarted apache using the WHM: httpd restarted

So how do I avoid this in the future? I decided to alter my domain template by removing the ssl log file. This file is unnecessary in my opinion – never use it, and doubt I will require it any other time.

vim /usr/local/cpanel/etc/httptemplates/apache1/ssldefault

Found the line for ssl_data_log and removed it.

cd /usr/local/apache/domlogs
rm *ssl_data_log

Restarted apache with peace of mind – of course, this is until WHM automatic update decides to rewrite my ssldefault file – but for now, I get to sleep comfortably.