[Pacemaker] process not getting started after a failure

[ESWAR RAO]

Hi All,

I have a 3 node cluster (node1,node2,node3).

The oc_pluginhandler resource is running in clone mode on 2 nodes as:
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
primitive oc_pluginhandler lsb:pluginhandler \
        meta allow-migrate="true" migration-threshold="10"
failure-timeout="300s" \
        op monitor interval="5s" timeout="120s"
clone oc_nvp_app_clone_pluginhandler oc_pluginhandler \
        meta clone-max="2" globally-unique="false" interleave="true"
location nvp_prefer_node_pluginhandler oc_nvp_app_clone_pluginhandler -inf:
node3
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Jul 17 16:32:12 node2 lrmd: [12219]: info: rsc:oc_pluginhandler:1
monitor[53] (pid 32144)
Jul 17 16:32:12 node2 lrmd: [12219]: info: operation monitor[53] on
oc_pluginhandler:1 for client 12222: pid 32144 exited with return code 7
(mapped from 3)

Jul 17 17:32:17 node2 lrmd: [12219]: info: rsc:oc_pluginhandler:1
monitor[53] (pid 12412)
Jul 17 17:32:17 node2 lrmd: [12219]: info: operation monitor[53] on
oc_pluginhandler:1 for client 12222: pid 12412 exited with return code 7
(mapped from 3)

Jul 17 18:32:19 node2 lrmd: [12219]: info: rsc:oc_pluginhandler:1
monitor[53] (pid 25174)
Jul 17 18:32:19 node2 lrmd: [12219]: info: operation monitor[53] on
oc_pluginhandler:1 for client 12222: pid 25174 exited with return code 7
(mapped from 3)


For some reason even though pluginhandler is not running on node2, the
node2 didn't intimate the  DC node(which was node1) .
So the process didn't get started on node2.
crm status shows that the process is running.

If I kill other processes on node2, then they are getting restarted
correctly.

Can someone please help how can I debug why that resource wasn't getting
restarted??????

Thanks
Eswar
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.clusterlabs.org/pipermail/pacemaker/attachments/20140722/cf9f9577/attachment-0002.html>