Disk Space Inaccuracy

[Matt Graham]

From: Brandon <brpacecar at gmail.com>
> We currently have a VPS server which runs in a Virtuozzo container.
> CentOS 5,  Apache 2.2.14
> /dev/vzfs              90G   78G   13G  87% /
> [root at production ~]# service httpd restart
> /dev/vzfs              90G   61G   30G  68% /
> However, a couple days later and I'm back in the same situation.

Log rotation not working right?  I've seen exactly these symptoms when
logrotate's been misconfigured.  You need to have logrotate do "apachectl
graceful" after it's rotated the logs; otherwise the httpds will have logfiles
still open even though logrotate's unlinked them.  Check that first.  You want
something like:

postrotate
 /usr/sbin/apachectl graceful

...in the right section of your /etc/logrotate.d/apache file; modify for your
setup.

Disk quotas should be totally orthogonal to this particular problem, unless
something really weird is going on.

-- 
Matt G / Dances With Crows
The Crow202 Blog:  http://crow202.org/wordpress/
There is no Darkness in Eternity/But only Light too dim for us to see