Link to home
Start Free TrialLog in
Avatar of ptsmartin
ptsmartin

asked on

Hyper-V Failover Cluster Disk Allocation

I'm setting up a failover cluster with 3 active nodes soon to be 4. And I'm running into a problem planning form my Hyper-V environment.

From: http://technet.microsoft.com/en-us/library/dd197460%28WS.10%29.aspx
MS recommends use a separate disk for each virtual machine: We strongly recommend that for each virtual machine, you use a separate disk (LUN) in cluster storage.

My question is, how do I create an environment with more than 20-23 (available drive letters) volumes for VMs?  Or is 20-23 (available drive letters) my limit?

Thanks
ASKER CERTIFIED SOLUTION
Avatar of msmamji
msmamji
Flag of Pakistan image
Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of ptsmartin
ptsmartin

ASKER

Great idea.  I'm using WS08 R2.
Avatar of ptsmartin
ptsmartin

ASKER

I have trouble moving VMs from one server to another when I use CSV.  If the VM host and the CSV Current Owner are not the same, the VM will not start.
I don't any trouble migrating VMs from one host to another if I have only 1 VM tied to a volume.
Avatar of msmamji
msmamji
Flag of Pakistan image
From What I understand this is what you are able to do.
http://blogs.technet.com/windowsserverexperts/archive/2010/01/29/hyper-v-csv-not-required-for-live-migration.aspx

So your live migration using CSV is not working. Let me know if you interested in fixing it.

Regards,
Shahid
Avatar of ptsmartin
ptsmartin

ASKER

Thanks for the link.  Here is the trouble I run into using CSV's.  They show as a placement for my VM in SCVMM, but when I select the drive letter and try to continue through a wizard (V2V or P2V), I get the following error:

Error (10608)
Cannot create a highly available virtual machine because Virtual Machine Manager could not locate or access C:\ClusterStorage\Volume1\.  

Recommended Action
Ensure that the path exists and that the path is for a cluster disk in available storage, and then try the operation again.

Do I need to set permissions on the CSV to allow SCVMM to write to it?  I do not have problems writing to individually assigned LUNs
Avatar of ptsmartin
ptsmartin

ASKER

I am able to place VMs on the CSV as long as the current owner of the CSV in the Cluster is the same as the server I'm placing the VM on.
Also I can can only migrate VMs from one host to another if the current owner of the CSV is the same as the new host.
Avatar of msmamji
msmamji
Flag of Pakistan image
Sorry for the late reply... was caught up in a few thing.

I will try to help in the new thread.
Regards,
Shahid.
Avatar of ptsmartin
ptsmartin

ASKER

I just deleted...My issue was a couple of things:
For the individual VM, my order for the Network for live migration was not in the correct order and enabled properly.  
This video helped solve that problem:
http://www.savilltech.com/Videos/2008R2Networks/2008R2Networks8006001mbps.wmv

And I did not have client for MS Networks selected on my Heartbeat Network which was being used as the live migration network.

Found that on this threat:
http://social.technet.microsoft.com/Forums/en/virtualmachinemgrclustering/thread/f60a4311-03a3-453f-891f-a1fa3d3f10d5

I'm going to migrate some VMs with unique LUNs to my CSV and see how things work.

Thanks again
Avatar of ptsmartin
ptsmartin

ASKER

I forgot to ask earlier.  How do you recommend disk allocation?
My CSV will be stored on a SAN.  Would it be best practice to allocate 1 drive that will accommodate all of my VMs and then a 2nd for snapshots?
Or would you recommend having several CSV volumes and distribute my VMs amongst them?  Same theory for the snapshots?

Thanks again