[Pacemaker] STONITH Monitor failure ... what should happen ?

[Andrew Beekhof]

On Thu, Sep 20, 2012 at 12:09 PM, David Morton <davidmorton78 at gmail.com> wrote:
> Simple question for today ...
>
> If a STONITH monitor fails, what is the designed behavior ? In our scenario
> we are using external/ipmi STONITH resources which interact with the IBM IMM
> (out of band management) controller.
>
> From what i can see at the moment its possible for the STONITH monitor to
> fail and not affect the cluster ?

Basically yes.

If you have a recurring monitor set up for it, we will notice and try
to recover it - at which point the device will either recover or be
marked as failed.
But it wont lead to the node being fenced at all (unless you set
on-fail= for the stonith operations).

> In saying that when this is the case the
> node which the STONITH resource is configured against is in standby and all
> resources are running on the node doing the monitoring ... if that makes
> sense ;)
>
> Does the fact that i put the node into standby then stopped openais cause
> pacemaker to not worry about the fact the monitor fails ?
>
> Am I missing something entirely or have misconfigured ?
>
> _______________________________________________
> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org
>