[Pacemaker] Recovering from failed stonith

Dejan Muhamedagic dejanmm at fastmail.fm
Tue Sep 8 15:32:56 UTC 2009


Hi,

On Fri, Sep 04, 2009 at 04:50:54PM +0200, Erik Hensema / HostingXS wrote:
> Op vrijdag 04 september 2009 14:41:18 schreef Johan Verrept:
> > On Fri, 2009-09-04 at 14:32 +0200, Florian Haas wrote:
> > > If it's not of any concern to you that your hard failover is not
> > > automatic, why not go with the meatware "plugin" (aka human
> > > intervention) in the first place?
> > 
> > Unless I misunderstood his setup, only when the power to the other half
> > the datacenter fails will the stonith fail. There are plenty of other
> > cases where the node seems dead but the stonith will work fine.
> 
> Indeed. I'm using stonith to recover from all the usual failures such as 
> kernel panics, resources refusing to stop, etc. However when the power of a 
> node fails human intervention is needed. I just didn't know how to do this, 
> but now I've found out. Thanks.

So which way did you adopt? One option is also to delete the node
which is definitely dead from the cluster configuration.

Thanks,

Dejan

> -- 
> Met vriendelijke groet,
> 
> 
> Erik Hensema / HostingXS Internet Services
> 
> _______________________________________________
> Pacemaker mailing list
> Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker




More information about the Pacemaker mailing list