[Pacemaker] on_fail

Andrew Beekhof beekhof at gmail.com
Fri Feb 13 10:29:26 UTC 2009


On Feb 13, 2009, at 11:23 AM, Romi Verma wrote:

>
>
> On Fri, Feb 13, 2009 at 3:36 PM, Andrew Beekhof <beekhof at gmail.com>  
> wrote:
>
> On Feb 13, 2009, at 11:05 AM, Romi Verma wrote:
>
> Hi All,
>
> pacemaker has one property "on_fail" for any operation configured  
> for a resource. suppose i have set "on_fail" to "fence"  for a  
> monitor operation , then is it expected to fence the node in case of  
> resource monitor failure ??
>
> that would seem like a reasonable assumption
>
> Thanks  Andrew for fast reply,
>
> i assume we need stonith for this.

definitely

> when resource monitor will fail on a node it means that node is  
> undergoing some problematic situation  ,  and due to this the  
> resource fails over to another node.
> Well , now  i am wondering  if a node is not well then stonith will  
> already come into the action.

if stonith is enabled (the default in 1.0.0 and onwards), then it  
kicks in whenever a node fails or a resource fails to stop.

> i mean to say even if i will not set "on_fail" to fence the node  
> will be fenced by stonith (as other members of cluster will detect  
> that node is not doing well and they will fence that errant node)
>
>  so why do we have this "on_fail" feature. is there any special  
> benefit of this "on_fail" feature i am missing??

setting on_fail=fence for a monitor op will cause the cluster to shoot  
the node immediately instead of trying to stop the resource and  
recover it without fencing.

>
>
>
>
> can any one explain please.
>
> Best Regards,
> _______________________________________________
> Pacemaker mailing list
> Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>
>
> _______________________________________________
> Pacemaker mailing list
> Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>
> _______________________________________________
> Pacemaker mailing list
> Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20090213/d18f5f47/attachment-0002.htm>


More information about the Pacemaker mailing list