Apache dies every Sunday, 6:00am
The Subject about says it all. For the life of me I can't track this
down. For some crazy reason, all apache processes cease to exist just
after (approx) 6:25am. This is the section of the error.logs:
[Sun Jan 5 06:25:28 2003] [notice] SIGUSR1 received. Doing graceful
restart
[Sun Jan 12 06:25:29 2003] [notice] SIGUSR1 received. Doing graceful
restart
And it has been doing this for quite some time. The only thing I can
tie it to is a cron job for logrotate. The apache settings have the
logs rotated on a weekly basis. I have tried to trace down WHY this
would happen, but came up with nothing. Here is the logrotate script:
/var/log/apache/*.log {
weekly
missingok
rotate 20
compress
delaycompress
notifempty
create 640 root
sharedscripts
postrotate
/etc/init.d/apache reload > /dev/null
endscript
}
I am using the sarge versions of both apache (1.3.26-1.1) and
logrotate (3.6.5-2). Any suggestions?
--
:wq
Matthew Daubenspeck
http://www.oddprocess.org
Reply to: