[Pacemaker] Pacemaker stops monitor on Master resource upon failovers

[Shyam]

Hi Folks,

I am using

pacemaker 1.1.6-2ubuntu3
heartbeat 1:3.0.5-3ubuntu2

that comes with ubuntu precise. For some reason the "monitor"
operation on master/slave resource is broken. It used to work in
previous versions. I enabled full debug logs & submitted a bug
http://bugs.clusterlabs.org/show_bug.cgi?id=5072 attaching the logs.

I tried to apply the LRMD patch in cluster-glue
http://hg.linux-ha.org/glue/rev/29c2eb9fa966 & also CRMd patch in
https://github.com/anchor/pacemaker/commit/3ab86ab9e8e63c920f7cf493f54eef7e8dfee32a

but nothing works. For some reason CRMD status appears to be bad for
the monitor despite the fact that it sends a cancel_op to lrmd.

Any help/pointers highly appreciated.

Thanks.

--Shyam