Dell openmanage nagios plugin is chatty
ianforde at gmail.com
Mon Jul 18 02:29:21 CDT 2011
On Thu, 2011-07-14 at 10:32 -0500, Eric Rostetter wrote:
> Quoting Jason Edgecombe <jason at rampaginggeek.com>:
> > the Dell openmanage plugin seems to give a warning on many problems that
> > are not important. The battery learning cycle is one such warning.
> Yes, annoying. On my monitoring, a battery learning cycle trips something
> like 4 or 5 warnings...
> > How
> > do other sites filter these non-critical events from the important events?
> We schedule our battery learning cycles for specific days (turn off the
> automatic running of them, then trigger them via cron). On that specific
> day, our monitoring staff gets the cron e-mail (so they remember that it is
> happening that day on that machine). They are also trained to know what a
> battery learn looks like in the monitor (so they don't ignore some other
> error as a battery learn issue). We then ack all the alerts except for
> the actual battery one, to cut down on the noise...
> I'd love a better solution than the above though. :) It is a real pain.
> Just knowing that the BB checks always happen on specific days cuts down
> on panic though. ;)
One thing you could do with this would be to have check_openmanage run
locally on each box from nrpe (rather than from the Nagios server,
assuming you're running Nagios) from a wrapper script. When the cron
job is triggered to start the battery learning cycle, a lockfile can be
set which the check_openmanage_wrapper script looks for. If the
lockfile isn't found, do the normal check. If it is found, do the check
with "-b bat_charge=ALL" instead. Oh - and make sure the cron removes
the lockfile at its end and on startup!
It's a big hackish, but you get the best of both worlds... I might play
with this idea myself at the office when I get the chance. Can you post
how you disabled automatic running and the cron job?
More information about the Linux-PowerEdge