S2D Cluster to Cluster CSV Test-SRTopology - can't get it to work

Gloria Gu 3,891 Reputation points
2020-07-16T06:10:04.403+00:00

Hello Technet,

I'm attempting to configure an S2D Cluster to Cluster and run the test-srtopology powershell command. I've run into a few snags along the way. I'm using the link below to configure.

https://learn.microsoft.com/en-us/windows-server/storage/storage-replica/cluster-to-cluster-storage-replication

I have configured the log and data volumes as CSVs which removes their drive lettering. I'm assuming I would then point the test-srtopology command at the c:\ClusterStorage\<volumename>. Unfortunately this fails. I've included the command I'm attempting below.

Test-SRTopology -SourceComputerName host1 -SourceVolumeName C:\ClusterStorage\VirtualMachines -SourceLogVolumeName C:\ClusterStorage\Logs -DestinationComputerName host2 -DestinationVolumeName C:\ClusterStorage\VirtualMachines -DestinationLogVolumeName C:\ClusterStorage\Logs -DurationInMinutes 5 -ResultPath c:\temp\
The error I get is:

Test-SRTopology : The specified volume C:\ClusterStorage\VirtualMachines cannot be found on computer host2. If this is a cluster node, the volume must be part of a role or CSV; volumes in Available Storage are not accessible.

The folders exist, I've disabled firewalls for tshooting, and I'm running in a powershell admin session.

What do I need to do to run the test with a CSV?

Thread source link: https://social.technet.microsoft.com/Forums/windowsserver/en-US/fb7013fc-5f08-4f31-b0ba-749dda4259e6/s2d-cluster-to-cluster-csv-testsrtopology-cant-get-it-to-work?forum=winserverClustering

Windows Server Clustering
Windows Server Clustering
Windows Server: A family of Microsoft server operating systems that support enterprise-level management, data storage, applications, and communications.Clustering: The grouping of multiple servers in a way that allows them to appear to be a single unit to client computers on a network. Clustering is a means of increasing network capacity, providing live backup in case one of the servers fails, and improving data security.
957 questions
0 comments
Accepted answer
  1. Xiaowei He 9,871 Reputation points
    2020-07-16T06:25:45.967+00:00

    Hi Andragoras,

    1. To enable storage replica, we just need to make the data disk to be CSV, the log disk shouldn't be configured to be CSV;
    2. It's not necessary to run the command "Test-SRTopology", it's just a test, and if we could enable the replication via GUI successfully, we may skip the command.
    3. To enable Storage Replica in Cluster, we may use the following step:

    a) Enable the Data disk in source node to be CSV

    b) Right click the CSV, Replication> Enable:

    c) Choose the related Data disk in replica site and choose the log disk in the wizard:

    Best Regards,

    Anne

1 additional answer

Sort by: Most helpful
Most helpful Newest Oldest
  1. hunter3740 96 Reputation points
    2020-11-26T05:04:57.483+00:00

    sticking with one pool (not a second pool for log vols), and will worry about crossing the 32 vol later (i.e. 32 data + 32 log = 64 max)

    just went for it:
    made a refs small log volume (on the available storage node, but none of the others), then the larger data vol as a clusterSharedVolume if you don't already have one (but move that to the ownerNode of available storage); repeat for other cluster; do the new partnership and watch the two refs log vols automatically get converted to cluster vols! 

    get-clusterResource 'available storage' | ft ownerNode
icm thatNode {new-volume -storagePoolFriendlyName myPool -friendlyName whateverLogs -fileSystem ReFS -storageTierFriendlyNames performance,capacity -storageTierSizes 36GB,4GB -driveLetter L}
# you probably have a data vol, but just e.g. new data vol
new-volume -friendlyName whatever -fileSystem CSVFS_ReFS -storagePoolFriendlyName myPool -storageTierFriendlyNames performance,capacity -storageTierSizes 100GB,900GB
get-clusterSharedVolume '*whomever*' | ft ownerNode
# *if* it's not "thatNode" (like available storage), move it there
# move-clusterSharedVolume -name "Cluster Virtual Disk (whatever)" -node thatNode
# repeat for other cluster the same way (other than it's node and server name will be different; for differentiation with below new-SRpartnership cmd, let's say the other cluster's node (that it's L is, in the owner of its available storage, and csv moved to that node are) is called "theOtherNodeCluster2"
# and also note, I ran this command from thisNode (on source computer):
new-SRpartnership -sourceComputerName thisNode -sourceRGname one-whatever -sourceVolumeName c:\clusterStorage\whatever -sourceLogVolumeName L:\ -destinationComputerName theOtherNodeCluster2 -destinationRGname two-whatever -destinationVolumeName c:\clusterStorage\whatever -destinationLogVolumeName L:\ -logSizeInBytes 36GB

    bam! the L: drive disappears on both cluster1 and cluster2
    side note, if you remove-srPartnership, you won't see L: drives come back yet, that will automatically happen when you "remove-SRgroup -name one-whatever" on cluster1 and "remove-SRgroup -name two-whatever" on cluster2

    so for sure, creating log vols as refs (not csvfs_refs like the data vols) works; i.e. they definitely get converted to csvfs_refs when you run new-SRpartnership

    assuming this only works because you stick to the recommended one pool (which again, where is the microsoft documentation that spells all this out?!)

    hope this helps someone