Apache (Occassionally) Won't Restart

[patrickkonsor]

Hello all,

When logrotate restarts Apache each night, it usually goes ok. However, occasionally a child process won't close and it just sits there doing nothing; I have to come in and manually kill -9 the hanging process. Here is an example of the logs..

[Wed Feb 11 01:10:01 2009] [notice] SIGHUP received. Attempting to restart
[Wed Feb 11 01:10:02 2009] [warn] RSA server certificate CommonName (CN) `localhost' does NOT match server name!?
[Wed Feb 11 01:10:02 2009] [notice] Apache/2.2.10 (Unix) mod_ssl/2.2.10 OpenSSL/0.9.8b DAV/2 PHP/5.2.6 configured -- resuming normal operations
[Wed Feb 11 01:11:04 2009] [warn] child process 14988 still did not exit, sending a SIGTERM

Is there anything I can do to prevent this from ever happening or to deal with it when it does happen so it doesn't just sit there until I notice something's wrong?

 

[scsi]

There has to be a reason its getting stuck. Maybe you will have to put apache into more of a debug mode so you can capture more then just a standard error like that. If not maybe you will have to change the way the init script restarts apache.

 

[RichardP]

Hello all,

When logrotate restarts Apache each night, it usually goes ok. However, occasionally a child process won't close and it just sits there doing nothing; I have to come in and manually kill -9 the hanging process. Here is an example of the logs..

[Wed Feb 11 01:10:01 2009] [notice] SIGHUP received. Attempting to restart
[Wed Feb 11 01:10:02 2009] [warn] RSA server certificate CommonName (CN) `localhost' does NOT match server name!?
[Wed Feb 11 01:10:02 2009] [notice] Apache/2.2.10 (Unix) mod_ssl/2.2.10 OpenSSL/0.9.8b DAV/2 PHP/5.2.6 configured -- resuming normal operations
[Wed Feb 11 01:11:04 2009] [warn] child process 14988 still did not exit, sending a SIGTERM

Is there anything I can do to prevent this from ever happening or to deal with it when it does happen so it doesn't just sit there until I notice something's wrong?

Same error here.
Still looking for the solve.

The result is that Apache is started by another user then root or apache. The ports 80 and 443 are locked by that other process, but Apache is not running correctly.

Only fix so far: lookup the processes with lsif -i :80 and do a kill -9 on them.

If you have the structural solve to prevent this issue, please mail me.