Some of you may have read from my previous Microsoft School Blog posts that I run a HyperV Clustered network. This consists of 3 HyperV Hosts, running Server 2008 R2 (considering when to upgrade to 2012)… with the VHD server data files located on a SAN.
This allows me to use the Failover Clustering feature to manage the resilience of the network.
This week; I came across an odd problem.
I was getting ready to do some maintenance across the system – and running a backup across the Hosts of each of the VMs located on them. As part of this process I wanted to move the owner of a Cluster Shared Volume (CSV) disk first. Cluster Shared Volumes are the bit of power that allows you to seemlessly present the same pool of storage with all the VHDs on to all the servers simulataneously. This make the live migration work – yes, thats the feature in VMware land called VMotion – that you pay for!
So, anyway – I digress… I go into Failover Manager find the CSV resource and select migrate to other node…. Only to be presented with
Operation has failed.
The action ‘Move to node <nodename>’ did not complete
Error code: 0×80071398. The operation failed because either the specified cluster node is not the owner of the group, or the node is not a possible owner of the group
What the!! Ive had enough problems with this cluster/SAN/CSV’s in the past – and its usually the DHCP virtual server cluster that refuse to connect their iSCSI disks…
So.. lets have a look at the “Possible Owners list” – and sure enough this server was missing a few command laster and were are back in business.
To tell who are the possible owners run:
Cluster res <resource name> /listowners
To to add a server to the list
Cluster res <resource name> /addowner :<servername>
Job done… oh, but remember, Server 2008 R2 needs you to run these commands from an elevated command prompt.