[Pacemaker] stop's on-fail of a master-slave resource is not effective

[Andrew Beekhof]

On Wed, Sep 19, 2012 at 7:58 PM, Kazunori INOUE
<inouekazu at intellilink.co.jp> wrote:
> Hi,
>
> In pacemaker-1.1,
> - CLusterLabs/pacemaker : b7db8b292d (Sep 18)
>
> on-fail of stop of a master-slave resource is not effective.
>
> [configuration of master-slave resource]
>  primitive prmAP ocf:pacemaker:Stateful \
>    op start interval="0s" on-fail="stop" \
>    op monitor interval="10s" on-fail="stop" role="Master" \
>    op monitor interval="20s" on-fail="stop" role="Slave" \
>    op promote interval="0s" on-fail="stop" \
>    op demote interval="0s" on-fail="block" \
>    op stop interval="0s" on-fail="block"
>  ms msAP prmAP
>
> Even if it specifies block as on-fail of stop,
> STONITH (fence) is performed at the time of stop failure.
>
> Is this a bug? Or specifications?

Bug, thanks for reporting.
Should be fixed now in:
   https://github.com/beekhof/pacemaker/commit/6a48a8b

>
> Best Regards,
> Kazunori INOUE
>
> _______________________________________________
> 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
>