Hi,
thanks a lot. As I only have max 10 servers on one virtualmin license its easy to keep track and check after a new server creation and correct if needed. Don’t want to bother them with this tiny issue, they have more to do.
Uwe
PS: for me thats solved, unless the server still crashes, but that would be a new thread
You should be aware that doing a graceful restart may take a long time to actually close the log files — read man apachectl for details.
I ran into the same Sunday Crashing situation. It all boils down to a few dozen restart requests in logrotate.conf that quickly follow one another:
‘postrotate
/etc/rc.d/init.d/httpd restart’
Once the amount of our virtual sites grew over ~20 we started experiencing this lovely phenomenon from time to time.
Frankly at this point I prefer the approach of processing log rotation and restarting Apache in order to avoid any crashes. Any downside to this approach? (info missing in the log files, etc?)