Bug 698890

Description of problem:
Freeze of vm based service is not viewable when using clustat

Version-Release number of selected component (if applicable):
3.0.12-10.el6.x86_64

How reproducible:
always

Steps to Reproduce:
1. have a vm based service running on node intrarhev1
[root@rhev1 ~]# clustat | grep rh56
 vm:rh56                        intrarhev1                     started       

From a libvirtd point of view:
[root@rhev1 ~]# virsh list 
 Id Name                 State
----------------------------------
 41 rh56                 running


2. freeze the vm based service with the domain running
[root@rhev1 ~]# clusvcadm -Z vm:rh56
Local machine freezing vm:rh56...Success

The server is still running from a libvirt point of view, as expected
[root@rhev1 ~]# virsh list 
 Id Name                 State
----------------------------------
 41 rh56                 running

3. run clustat again
  
Actual results:
[root@rhev1 ~]# clustat | grep rh56
 vm:rh56                        intrarhev1                     started   

Expected results:
[root@rhev1 ~]# clustat | grep rh56
 vm:rh56                        intrarhev1                     started   [Z]

Additional info:

The service is indeed frozen, in the sense that if you try for example to stop it you get an error:
[root@rhev1 ~]# clusvcadm -s vm:rh56
Local machine stopping vm:rh56...Failure: Service is frozen

Also, if you stop the guest with virsh command
[root@rhev1 ~]# virsh shutdown rh56
Domain rh56 is being shutdown

it is not recovered as it is indeed frozen, so not monitored by rgmanager.
With the domain yet stopped, if you run
[root@rhev1 ~]# clusvcadm -U vm:rh56
Local machine unfreezing vm:rh56...Success

the domain is then correctly restarted by rgmanager.
So all is good from a functional point of view, but it would be nice to have clustat give a notice of the frozen status of the service, as it is for "conventional" services.

Gianluca