Microsoft Enterprise Platforms Support: Windows Server Core Team
EPS Team Blogs
Product Team Blogs
One question I get asked a lot working with our customers is how to failover the ‘Cluster Group’ and ‘Available Storage’ groups?
Under the covers, those two groups are still just groups like any other you’d see under ‘Services and Applications’ in Failover Cluster Manager. We just didn’t expose through the GUI, a way to fail those groups between nodes when you want to move say ‘Available Storage’ to a node you are working on so you can copy data to a cluster shared disk.
The resources in the ‘Cluster Group’ are located in the middle pane in Failover Cluster Manager when you highlight the name of the cluster underneath ‘Failover Cluster Manager’ in the left-hand pane.
Note in the above screenshot, in the top ‘Summary of Cluster’ box, the ‘Current Host Server: shows as JH-W2K8-R2SP1-2. That’s the node that currently owns the ‘Cluster Group’ or ‘Core Cluster Resources.
The below screenshot shows the ‘Available Storage’ group. This is a group containing cluster shared disks that have not yet been added to a ‘Services and applications’ group.
Note also that the current owner of the ‘Available Storage’ group shows as JH-W2K8-R2SP1-2.
So, there are two ways we can move either the ‘Cluster Group’ or the ‘Available Storage’ groups between nodes of the cluster.
The first way is with the cluster PowerShell cmdlet Move-ClusterGroup
Move-ClusterGroup “Cluster Group”
Move-ClusterGroup “Available Storage”
The second way is a cluster.exe command line
cluster group “Cluster Group” /move
cluster group “Available Storage” /move
If you have more than two nodes in the cluster, you can specify the destination node with the following syntax:
move-ClusterGroup –node NodeName
where NodeName is where you want to move the group.
cluster group “Cluster Group” /Move:NodeName
Jeff Hughes Senior Support Escalation Engineer Microsoft Enterprise Platforms Support
The same approach applies to Cluster Shared Volumes (CSVs). Each of them is essentially a hidden cluster group as well.
Currently, neither the group nor resource names for CSVs are exposed in PowerShell when listing. Cluster.exe doesn't show the group as well. So the only way to learn the group name is to list the resources. You filter the list by the friendly name of the CSV or a part of it (as seen in Failover Cluster Manager).
PS C:\Windows\system32> .\cluster.exe /cluster:"MyCluster.contoso.com" resource | findstr /i Disk-01
Cluster Disk 1 - Disk-01 53f2937f-feb2-4dd4-ac20-7e2531023455 Node02 Online
The GUID is actually the Group name. (And it is not the same as internal object GUID). Knowing this name you can now use it either with Cluster.exe or with PowerShell.
Get-Cluster -Name "MyCluster.contoso.com" | Get-ClusterGroup -Name "53f2937f-feb2-4dd4-ac20-7e2531023455" | Move-ClusterGroup
Thanks, nice bookmark. :-)
This is incorrect. We are running Exchange 2010. I just attempted to use this command. We have 6 servers in a cluster group named "cluster group". The server I intend to host the cluster service is eupdcmail003p. I launch a dos prompt, and initiated this command.... cluster group “Cluster Group” /Move:eupdcmail006p I receive the following error. "Too many command line parameters have been specified for this option.
I'm curious. I see Available Storage when I list the cluster groups via powershell (get-clustergroup) but when I look at Failover Cluster Manger there is no group listed in the Storage Summary. Also, the Storage Summary header lists Available Capacity as: Total: 0 Bytes, Free Space: 0 Bytes, Percent Free: 0%
Why is there an Available Storage group if there is no available capacity?
Is ther any different in using the GUI and Powershell and the Cluster.exe command ?
This article helped me solve a problem bringing online some disks that were not yet a part of a group in the cluster. I'm setting up a "geo-cluster" and these disks would try to come online on the node they haven't been added to yet. I thought it was a problem with the disks themselves, rather than an ownership issue with the "Available Storage" group. It all makes sense now.
Hi, do you need to run cluster validation after moving the cluster group?
Thanks for the info presented here,
No, you do not have to run Cluster Validation if you move the Cluster Group to another node
From command line:
cluster "clustername" group "groupname" /move:nodename
Cluster name needs to be mentioned; in case; it says that syntax is not correct.
WHY didn't you put it in the GUI? It was there in all previous versions of the OS that supported clustering. This sort of stuff is just frustrating...
In 2012 and above, it was added to the GUI so you cam do failovers of these two groups more easily.