[Mailman-Users] Old problem, AGAIN!

[Brian Teal]

OK, this happened to me a while ago and I've included some of your previous
answers at the bottom of this post, but here is the problem.
I run a hosting company that is subcontracting the servers we use from
another larger company (resellers, yes). They are not quite "dedicated" but
we do have our own processing, resources, etc. on the servers so when we
restart our "server" it isn't necessarily restarting the whole physical
server (UNIX running Red Hat). Anyway, one of the servers had a "physical
fault" today and had to be restarted by the technicians. After it was
restarted, mailman was no longer sending out messages. I tailed the messages
into mailman, into the "wrapper"? But nothing comes out to my 350
recipients. I checked the crontab and looked for qrunner somewhere, but this
is what it looks like:
___________________________________
SHELL=/bin/bash
PATH=/sbin:/bin:/usr/sbin:/usr/bin
MAILTO=root
HOME=/

# run-parts
* * * * * root run-parts /etc/cron.minute
*/10 * * * * root run-parts /etc/cron.tenth
01 * * * * root run-parts /etc/cron.hourly
02 4 * * * root run-parts /etc/cron.daily
22 4 * * 0 root run-parts /etc/cron.weekly
42 4 1 * * root run-parts /etc/cron.monthly
___________________________________

It doesn't contain qrunner anywhere in it, should it?
Also, when tailing the message, it goes to
/home/mailman/mail/mailman_wrapper post cbr900 and it appears as sent.
However, nothing comes out after that.
So, with all of that, here are the other posts from you all last time I had
this problem. Let me know what you think. I HAVE done the killall, checked
perms and db and all are fine. I reset the cron and the system time looks
correct.

*****<<<SNIP>>>*****
On Wed, Jan 16, 2002 at 08:01:35AM -0500, Barry A. Warsaw wrote:
>
> >>>>> "DW" == Dan Wilder <wilder at eskimo.com> writes:
>
>     DW> If you can get to the mail logs on the server (someplace in
>     DW> /var/log/ maybe, often called "mail." something, and do a
>     DW> "tail -f logfilename" while you send some mail to the list,
>     DW> you can sometimes gain some insights.
>
> Yep, also check the Mailman logs in $prefix/logs and check the Mailman
> queue $prefix/qfiles.  What version of Mailman?  What version of
> Python?  Did any of these change?  Did they upgrade any system
> software?  Did they muck with cron?  Maybe your crontab entries got
> wasted.
>
> >>>>> "PTS" == Pro-phile Technology Solutions <mailman at pro-phile.com>
writes:
>
>     PTS> I recently upgraded my server space (about 2:30 today) and
>     PTS> the list stopped working at about 2:30 today. Hmmmm. They
>     PTS> (the techs) said that the only thing they did was allocate
>     PTS> more space on the servers I was on for me to use and they
>     PTS> touched nothing else.
>
> As a former (and sometimes current reluctant) sysadmin, I just don't
> believe this. :)  It's too suspicious.

As a current sysadmin, amen!

The original querant had emailed me personally and I replied
without thinking to copy the list.  I'd suggested he check
cron, both the crontab (or /etc/cron*/whatever) entry for
qrunner, to see whether it had been deleted, and the cron daemon.
Sometimes after a severe clock-shaking it's useful to
"killall -HUP crond" to get it moving again.

Of course the prudent sysadmin would reboot at that point,
as crond isn't the only daemon that'll sulk following
a period of temporal chaos.  Postfix is my favorite; short of
reboot, I'll always do a "postfix reload" following a significant
clock jump on a server I can't reboot at that time.

--
-----------------------------------------------------------------
 Dan Wilder <dan at ssc.com>   Technical Manager & Editor
 SSC, Inc. P.O. Box 55549   Phone:  206-782-8808
 Seattle, WA  98155-0549    URL http://embedded.linuxjournal.com/
-----------------------------------------------------------------