On Thu, 3 Feb 2000, Ounsted, Toby wrote:
> My named is dying! Here's what I'm seeing: named chugging
gracefully along
> doing its stuff every hour and responding to queries. The
only thing going
> on in between is SYSLOGD RESTART - after which named takes
no further
> interest in doing anything until I reset named (KILL -HUP)
at 9:42 at which
> point everything is happy again. Has anyone else
experienced similar
> behaviour?? (I think) I've now seen this behaviour 2
Mondays in a row..
>
> It's a stock 6.1 installation with no component updates.
If it's happening
> to me then it's got to be happening somewhere else....
How are you watching the log? If it's with "tail -f" then
chances are the
name server IS still doing it's thing, but it's being logged
to a file
other than the one you're watching. Try restarting your
"tail" after
syslog restarts, you'll probably see it there.
The syslog restart, by the way, is most likely just a sign
that logrotate
is doing it's thing. You'll probably find a logrotate
script in your
/etc/cron.daily and "killall -HUP syslogd" in your
/etc/logrotate.conf
Cheers.
I'm pretty sure it's the logrotate - the presumption being
that named is trying to write to the file, the file gets rotated, named
commits suicide. The obvious one is to have named log to a file other than
'messages' and ot rotate that file but it's not a real fix and I don't think
that named should behave that way in the first place..
T.
--
To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe"
as the Subject.
--
To unsubscribe: mail [EMAIL PROTECTED] with "unsubscribe"
as the Subject.