[Pacemaker] Avoiding an email flood with MailTo

[Andrew Beekhof]

On Fri, Aug 14, 2009 at 11:18 PM, Dan Urist<durist at ucar.edu> wrote:
> I have a 2-node cluster which runs several sets of vservers. Each set of
> vservers is dependent on a different drbd volume. I would like to use
> the MailTo agent to notify me when a resource fails over (either by
> colocating an instance of the MailTo RA with each vserver and drbd
> volume or including them together in a group), but if possible I would
> like to constrain the MailTo agent so that if a drbd volume fails over,
> I only get an email for that event and not a flood of emails for each
> vserver that depends on that volume.
>
> From what I've seen of the behavior of the MailTo agent, I don't think
> this is possible, since it sends an email when it starts and stops and
> it would stop when the vserver stops.

Depends where you have it in the group.
(I assume its in a group?)

If you have it at the end, it will behave as you described.
If you have it directly after the drbd resource, it will behave as you wanted.

> I could modify the MailTo RA so it
> only sends an email when it starts, but then I wouldn't receive an email
> if the vserver couldn't be started (which of course I want to know).
>
> Is there a way to do this that I'm missing?
> --
> Dan Urist
> durist at ucar.edu
> 303-497-2459 (office)
> 303-961-2675 (cell)
>
> _______________________________________________
> Pacemaker mailing list
> Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>