Apache crashes; error message on restart

[Tipem]

Hi,

Occasionally Apache will crash on me, when our server is under a completely normal load. When I run /etc/init.d/httpd restart, this is the error I get (it fails on restart):

Stopping httpd: [FAILED]
Flushing IPC Semaphores [ OK ]
Starting httpd: (98)Address already in use: make_sock: could not bind to address [::]:80
(98)Address already in use: make_sock: could not bind to address 0.0.0.0:80
no listening sockets available, shutting down
Unable to open logs
[FAILED]

What causes this kind of error? It's totally bizarre to me and nothing has changed in our configuration; this error just started popping up when I try to restart Apache after it crashes.

Any ideas? Thanks in advanced. This place is very helpful! :cool:

Ian

[falko]

I guess that there's still an Apache zombie process running on port 80 when you try to start Apache.
Are there any errors in Apache's error log?

[Tipem]

I am getting a crapload of segmentation faults (11).

e.x.,

[Tue Sep 01 17:53:29 2009] [notice] child pid 6131 exit signal Segmentation fault (11)
[Tue Sep 01 18:13:06 2009] [notice] child pid 11996 exit signal Segmentation fault (11)
[Tue Sep 01 18:19:01 2009] [notice] child pid 13249 exit signal Segmentation fault (11)

etc, etc etc. Like every line in my error log for the past 3 days is this error. What does this mean? Any ideas?

I sincerely appreciate your help.

[falko]

This happens from time to time, even on my servers. I'm using monit to scan the logs for this and restart apache when this happens.

What's the output of netstat -tap?

[Tipem]

I ran the command and the list hasn't even finished loading after several minutes. There are HUNDRED of lines to local address "http" and the state is all "TIME_WAIT". The pid to all of these TIME_WAIT processes appears as "-".

Every several lines, the line getnameinfo failed appears.

Any ideas?

[falko]

What's in /etc/resolv.conf?
Have you tried to reboot the server?