[Mailman-Users] mailman watchdog - status report

[Matt Emerson]

On Dec 21, 2005, at 4:02 PM, Con Wieland wrote:

> I recently had some (more) problem's that Mark Shapiro helped me out
> with. In that exchange I asked about a watchdog type program to help
> in early problem detection.

One thing we use here is a script that keeps an eye on the qrunners.

They've died on us for some reason once or twice, so we use this  
script to notify us of the problem.  We run it from cron.

#!/bin/sh
# check-qrunner
# see if mailman's qrunners appear to be running and squawk if
# they're not.

if ps axw | grep -v grep | grep -v check-qrunner | grep qrunner >/dev/ 
null; then :; # life is good
else
         # life is bad.  try to restart the qrunners, and tell someone
         /usr/share/mailman/bin/mailmanctl -s start
         worked="did not work";
         if ps axw | grep -v grep | grep qrunner >/dev/null; then
                 worked="worked";
         fi
         echo "Hi.  I am /usr/local/bin/check-qrunners on `hostname`.

The Mailman qrunners seem to have died.  It appears that my
attempt to restart them $worked." | mail -s "mailman qrunners"  
someone at nasa.gov
fi



-- 
Matt Emerson
rme at grc.nasa.gov