Apache and Named Crashing

[defamai]

Hello,
Apache and named seem to be crashing out of the blue, for no apparent reason, several times per day.
I see the following in the logs when this occurs:
[Sat Apr 15 16:01:00 2006] [notice] caught SIGTERM, shutting down

I've run httpd configtest, and recieve no errors.

s1# httpd -v
Server version: Apache/1.3.34 (Unix)
Server built: Apr 14 2006 00:52:52

Any help resolving this would be much appreciated.

 

[DirectAdmin Support]

Hello,

Sigterm is the standard "shutdown" signal.
If there is nothing in the /var/log/httpd/error_log just after the sigterm, then might be one of a few things:

1) full disk
2) oversized error_log
3) too many filedescriptors: http://help.directadmin.com/item.php?id=9

John

 

[defamai]

Hello,

After debugging, we traced the problem to direct admin inserting this into its task.queue file, which was then being processed by dataskq:
"action=named&value=reload"
I inserted debugging lines into /usr/local/etc/rc.d/named, so that it outputted text to /root/named_log, and found this:

s1# cat /root/named_log
Mon Apr 17 03:10:00 EDT 2006
Stopping Named: [ OK ]
Starting Named:RETVAL 127
[ FAILED ]
Mon Apr 17 03:10:01 EDT 2006
Stopping Named: [ FAILED ]
Starting Named:RETVAL 127
[ FAILED ]
Mon Apr 17 03:10:02 EDT 2006
Starting Named:RETVAL 127
[ FAILED ]
s1#

I'm not sure whats going on, and all help is much appreciated.

 

[defamai]

Well, we found the solution.

Changed
daemon /usr/sbin/named -u bind
to
/usr/sbin/daemon /usr/sbin/named -u bind
in /usr/local/etc/rc.d/named
and, did something simailar in /usr/local/etc/rc.d/httpd
Thank you for your time.