Monitoring PE2950 Status with ESXi?

Steve Jenkins stevejenkins at gmail.com
Wed Jul 13 10:15:17 CDT 2011


On Wed, Jul 13, 2011 at 6:32 AM, Mark Rolen <mrolen at opubco.com> wrote:
> We're having very good luck with the check_esxi_hardware.py plugin for
> Nagios (uses WBEM, so nothing to install on ESXi).  Gives you alerts on
> fans, disk, power supplies, etc.  It just alerted us on a predictive
> failure for a disk in one of our R900s yesterday.
>
> Regards,
> Mark

Thanks for everyone's excellent feedback. We've grown very accustomed
to having Nagios monitor everything through the check_openmanage
plugin, which does EXACTLY what we want and has saved our bacon a few
times with failing DIMMs, early drive and array failure predictions,
etc. We're also very happy with how we've configured Nagios to only
page us in the middle of the night for the IMPORTANT stuff, while the
minor stuff (like a 20% disk space remaining warning) can wait until
"work hours," and generally just tweaked our monitoring platform to
our liking, so I was really looking for something that will provide
very similar functionality without having to relearn or redeploy a
bunch of stuff.

The check_esxi_hardware plugin looks promising. I'll definitely check
that out. As a worst case, my thought is that we could run a minimal
install CentOS 6 server as VBox or VMWare host, run nothing but OMSA
and the Nagios NRPE interface on the host server, then launch all our
guests from that platform.

In general, those using VMWare (ESX or ESXi) - are you happy with
performance and compatibility on the PowerEdges? I've been
experimenting with virtualization for the past few months using Oracle
VirtualBox as the host system on a number of different PowerEdges -
including a dinosaur 2450 (which actually worked... VERY slowly). But
for a production environment, we're thinking we should go with
something "industrial strength."

SteveJ



More information about the Linux-PowerEdge mailing list