One of my servers started groaning early last week and it turned out I had log files accumulating to nearly 1TB. Log rotation was switched on but it still wasn’t ideal.
I solved this by ssh’ing into a root shell and running the following command:
du -a /var | sort -n -r | head -n 10
This lists the directories which are taking up the most space and it showed up the locations of the log files, for me using plesk and ubuntu they were:
/var/www/vhost/domainname.com/statistics/logs
rm *
Removed them but then I went on to stop it writing to the access log. This can be done by deleting all lines which mention access_log in:
/var/www/vhost/domainname.com/conf/httpd.include (or whatever the file is called, one server had a random number as the filename.
/etc/apache2/httpd.conf
to delete the lines just nano /etc/apache2/httpd.conf hold down ctrl w and search for access_log then I think it was ctrl k to delete the line. ctrl x writes out and quits. Then finally restart apache
/etc/init.d/apache2 restart
Server log files shouldn’t cause a problem again.
Disk space usage went from 80% to 1% and Ram usage went from 75% to 45%. Faster server = better conversion rate = improved ROI.
