If you have two Hyper-V failover clusters managed by System Center Virtual Machine Manager 2008 R2, it is possible that you can't complete a LAN migration of a Virtual Machine (VM) from one of the clusters to the other. This issue will only appear if ALL of the following conditions are met.
- VM is created using Hyper-V (i.e. the VM was not created through the VMM Administrator Console)
- VM is made into an HA VM using the FOC GUI (resource name will be called something like “Virtual Machine <vmname>”)
- VM is in a running state
- VM is migrated to a different cluster. VMM will chose QSM (Quick Storage Migration) network migration (LAN migration) in this case
When the migration fails you will see the following error:
VMM cannot complete the WMI operation on server <serverName> because of error: [MSCluster_ResourceGroup.Name="2068e895-4930-42be-a4c8-152ab15a28b8"] The cluster group could not be found.
(The cluster group could not be found (0x1395))
Resolve the issue and then try the operation again.
1. If the VM is in a saved state and not in a running state, at the beginning of the migration across the two clusters then this issue will not appear and everything completes successfully. Due to this, one resolution is to put the VM in a saved state first and then initiate the migration.
2. Use the Failover Cluster Manager User Interface to locate the Virtual Machine in "Services and Applications". Right click on the top resource group for this VM and change the Resource Name from "Virtual Machine <vmname>" to "SCVMM <vmname>". Now refresh this VM from VMM using the "refresh-vm -force '<vmname>'" cmdlet. Migrating the VM from one cluster to another should now complete successfully.
If you are already in a situation where your VM migration failed (the virtual machine should be in a migration failed state) there is only one way to get your VM in a healthy state and that is by following the steps below:
1. Take a full database backup of the DB used by VMM (Just in case; this is a safety net in case something goes wrong)
2. Use the instructions in the following blog post to save any metadata applied to the VMs in your cluster. http://blogs.technet.com/b/m2/archive/2010/04/16/saving-and-re-applying-the-virtual-machine-metadata-in-vmm.aspx
3. Remove the source cluster from management. Here we assume that the Migration Failed VM is part of this source cluster (i.e. the cluster that contains the host that had the VM in the first place before the migration was initiated)
4. Re-add the source cluster into SCVMM management. The VM in question should now be in a healthy state and running. Verify your VM is working properly in the source cluster. This is a very important step.
5. Go to the destination cluster (the cluster that contains the host you want to move the VM to) and cleanup the service that maps to this VM. You might also have to go to the destination hyper-v host and delete that VM that was created as a placeholder. This is a very important step. You need to ensure you are deleting the VM from the destination cluster [the VM should be working properly in the source cluster - DO NOT touch that]
6. Re-apply any metadata to the VMs of your source cluster after it was re-added into management by SCVMM. The blog post referenced above explains how to do that.
7. Refresh the source and destination clusters.
8. Refresh the Virtual Machine that is going to be migrated.
9. Start the migration of the VM from the source cluster to the destination cluster.