Partager via

How to Failover the ‘Cluster Group’ and ‘Available Storage” Cluster Groups

  • Article

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.

FCM

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.

AvailableStorage

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”

MoveGroupPS

or

Move-ClusterGroup “Available Storage”

MoveAvailableStoragePS

The second way is a cluster.exe command line

cluster group “Cluster Group” /move

MoveGroup

or

cluster group “Available Storage” /move

MoveAvailableStorage

If you have more than two nodes in the cluster, you can specify the destination node with the following syntax:

PowerShell:

move-ClusterGroup –node NodeName

where NodeName is where you want to move the group.

Command Line:

cluster group “Cluster Group” /Move:NodeName

where NodeName is where you want to move the group.

Jeff Hughes
Senior Support Escalation Engineer
Microsoft Enterprise Platforms Support

Comments

  • Anonymous
    January 01, 2003
    @Marios Philippopoulos:
    No, you do not have to run Cluster Validation if you move the Cluster Group to another node
  • Anonymous
    January 01, 2003
    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:Windowssystem32> .cluster.exe /cluster:"MyCluster.contoso.com" resource | findstr /i Disk-01Cluster Disk 1 - Disk-01 53f2937f-feb2-4dd4-ac20-7e2531023455 Node02    OnlineThe 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
  • Anonymous
    January 01, 2003
    @Ken
    In 2012 and above, it was added to the GUI so you cam do failovers of these two groups more easily.
  • Anonymous
    February 04, 2012
    Thanks, nice bookmark. :-)
  • Anonymous
    June 15, 2013
    The comment has been removed
    • Anonymous
      April 07, 2016
      You cannot copy and paste the command and expect it to work. Oddly enough, I deleted both of the " and typed it out manually and the command worked. Guess copy and paste is not always the best. :)
      • Anonymous
        October 05, 2017
        why does paste not work, it's all text so why does this not work unless you type this out?
  • Anonymous
    July 09, 2013
    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?Thanks
  • Anonymous
    December 04, 2013
    Is ther any different in using the GUI and Powershell and the Cluster.exe command ?
  • Anonymous
    December 13, 2013
    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....thanks!
  • Anonymous
    November 29, 2014
    Hi, do you need to run cluster validation after moving the cluster group?

    Thanks for the info presented here,
    Marios Philippopoulos
  • Anonymous
    December 12, 2014
    The comment has been removed
  • Anonymous
    February 21, 2015
    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...
  • Anonymous
    August 02, 2015
    Hello Jeff,

    Kindly share the step by step document for Cluster setup. I want to setup the SQL Cluster.
    Email Address: srtavish@gmail.com

    Thanks,
  • Anonymous
    January 22, 2016
    Hi All, Greetings for the day!

    I have an issue- There is a need to use clustered nodes for my application because of volume. A 3rd party vendor did clustering for us. First they had some issues using Quorum for sharing application resources. So, they have created a clustered disk which uses our App resources. So now we have one Active and one Passive node for our application.

    Now, whenever user is running his session, it remains active and does not get disturbed on switching nodes from Active to Passive. But, it becomes inactive and of no use when switching from Passive to Active.

    We ran the Failover Cluster Validation. In the wizard 'Review Storage Status' , I selected the clustered disk and not the Quorum volume. Here are the category results:

    Cluster Configuration - Success
    Inventory - Success
    Network - Warning
    Storage - Warning
    System Configuration - Failed

    Now, here are the results for 'System Configuration' :
    Validate Active Directory Configuration - Failed
    Validate All Drivers Signed - Success
    Validate Cluster Service and Driver Settings - Success
    Validate Memory Dump Settings - Success
    Validate Operating System Edition - Success
    Validate Operating System Installation Option - Success
    Validate Operating System Version - Success
    Validate Required Services - Success
    Validate Same Processor Architecture - Success
    Validate Service Pack Levels - Success
    Validate Software Update Levels - Success
    Validate System Drive Variable - Success

    The Overall Result is:
    Overall Result
    Testing has completed, but one or more tests indicate that the configuration is not suitable for clustering.

    Details for Failed section 'Validate Active Directory Configuration' is as below:

    The distinguished name of node A could not be determined because of this error: There was an error getting information about the organization unit for node 'A' from the domain 'company.com'.
    The organizational unit of node A could not be determined because of this error: Did not find an Organization Unit (OU) in the Active Directory
    The distinguished name of node JDAPRDWMPS.sophiemartin.com could not be determined because of this error: There was an error getting information about the organization unit for node 'JDAPRDWMPS.sophiemartin.com' from the domain 'company.com'.
    The organizational unit of node A could not be determined because of this error: Did not find an Organization Unit (OU) in the Active Directory
    Connectivity to a writable domain controller from node A could not be determined because of this error: Could not get domain controller name from machine.
    Node(s) JDAPRDWMAC.sophiemartin.com can reach a writable domain controller.
    Node(s) JDAPRDWMPS.sophiemartin.com cannot reach a writable domain controller. Please check connectivity of these nodes to the domain controllers.
  • Anonymous
    October 01, 2016
    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?
    • Anonymous
      October 04, 2016
      The Available Storage group will always be there whether there are any disks in it or not. When you first add a disk to Failover Cluster, it goes into the Available Storage group. From there, you can then add it to an Application Group such as SQL or a File Server.
  • Anonymous
    June 02, 2017
    Thanks. What I'm trying to do is create scripts to move ALL resources to 1 node so that node can be patched, then after patching run the same script on the other node to move all resources to the 1st node that was patched. Then I'll need another script to manually rebalance the cluster since it's active-active.