I have a Hyper-V cluster (2008 R2) with a CSV that I want to clone. The CSV is on a fibre channel SAN (HP MSA M3 P2000 G3). I cloned a volume from the SAN side of things, and have mapped it to my cluster nodes. The cluster nodes can see the volume, but they are marked as offline, reserved.
As far as I can tell, I am having the same problem as outlined in this question: Changing 'disk ID' of Windows GPT disk via Linux except that I'm working with fibre channel.
When I run
diskpart and get the
uniqueid disk, both the original disk and cloned disk share the same ID. I think this is the problem.
DISKPART> select disk 10 Disk 10 is now the selected disk. DISKPART> uniqueid disk Disk ID: 65DC665F DISKPART> select disk 4 Disk 4 is now the selected disk. DISKPART> uniqueid disk Disk ID: 65DC665F
Additionally, when I check the attributes on the clone, it shows as clustered:
DISKPART> attributes disk Current Read-only State : Yes Read-only : Yes Boot Disk : No Pagefile Disk : No Hibernation File Disk : No Crashdump Disk : No Clustered Disk : Yes
I've also checked on the SAN, and there are no reservations on the cloned volume.
The solutions I've running across so far:
Online the disk from the cluster owner of the original volume, at which point the conflicting unique ID should be resolved. This is not working for me, I am unable to online the disk from the CSV's current owner.
Map the volume to a non-clustered server, which should be able to mount the volume, and then modify the uniqueid. Which I can't do, only cluster nodes are plugged into the FC switch.
Offline the original CSV, bring the clone online. An option I'd like to avoid, the CSV contains a production server's VHD.