> > I found the following stonithd behavior.
> > It might be an expected one, but I'm just wondering.
> >
> > My operation is here;
> > (1) start Heartbeat 2.1.4 on two nodes(dom-d1, dom-2).
> > (2) start the resource on active node(dom-d2), and dom-d2 is also DC in
this
> > case.
> > (3) modify the RA and cause stop NG on dom-d2.
> > (4) standby node(dom-d1) would STONITH to dom-d2.
> >    stonith plugin is xen0.
> > (5) STONITH action succeeded.
> >
> > When I "grep" the ha-debug on dom-d1(previous standby node) after
STONITH,
> > I found that there are duplicate "reset" actions like this;
> > Is this expected?
> 
> For the DC, basically yes.
> The old DC fences itself but obviously isn't around to record that it
succeeded.
> The new DC just sees that the old one is in a strange state and the
> only safe thing to do is shoot it (again).

I see, make sense.
By the way, Pacemaker will do the same in this case, right?

Thanks,
Junko

_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to