HA: strange behavior

[mir]

Hi all,

On one node HA has simply stopped working. Removing the node and I can start it again but as soon as the vm is added to HA again I see the following in rgmanager.log:
Sep 24 22:37:06 rgmanager Initializing pvevm:115
Sep 24 22:37:06 rgmanager pvevm:115 was added to the config, but I am not initializing it.

A manual restart logs this:
Executing HA start for VM 115
Member esx2 trying to enable pvevm:115...Aborted; service failed
TASK ERROR: command 'clusvcadm -e pvevm:115 -m esx2' failed: exit code 254

And rgmanager.log:
Sep 24 22:48:40 rgmanager #43: Service pvevm:115 has failed; can not start.
Sep 24 22:48:40 rgmanager #13: Service pvevm:115 failed to stop cleanly


What could be wrong here?

 

[dietmar]

Can you start/stop the VM when HA is disabled?

 

[mir]

Can you start/stop the VM when HA is disabled?

Yes, this goes for all vm's on the same node.

 

[dietmar]

And you disabled the service before re-enabling it?

 

[mir]

And you disabled the service before re-enabling it?

Yes. I have even tried to reboot the node but the problem persists. Migrating HA enabled vm from another node to this node also works as expected. If I migrate offline the vm's from this node to another node and try to enable HA for these vm's on another node the server error is seen so the problem is to be found in cluster wide config/settings.

The node is correctly registered in the cluster and all cluster functionality also works on the node.

 

[mir]

After upgrading to latest version yesterday this problem vanished. Since I don't know whether this issue was I bug in the previous version which was solved with yesterday's backup or resolved some other way I will not mark the issue as solved.