Volume move to new aggregate and Snapmirror

[BenBrazil]

Hi,

We've got two clusters in separate Data Centres. Cluster 1 from a snapmirror perspective is our source and Cluster 2 is our destination. We use snapmirror on the volume and not on the SVM.

We've just added two new nodes in each cluster and therefore two new aggregates. We will be decommissioning two nodes in each cluster and will therefore need to move the volumes to the new aggregates on the new nodes. The volumes will be staying on the same SVM in each cluster.

Is the process simply to quiesce the snapmirror, perform the volume move on the source and destination and then resume the snapmirror, or do I need to delete the snapmirror relationship and start again?

Or should I delete the snapmirror destination volume, create a new one and then create a new snapmirror and initialize?

Is there anything else that I am missing?

Thanks,

Ben

[RvanDantzigITQ]

You can use the volume move command in ONTAP to move a SnapMirrored volume, whether it is a source or destination volume. The process is nondisruptive and does not require breaking the SnapMirror relationship. During the move operation, SnapMirror updates may briefly pause, but they resume automatically after the cutover phase is complete.
This functionality is supported starting from ONTAP 9 and allows seamless migration of SnapMirror volumes to another aggregate within the same cluster

[Sergey_Osipov · Accepted solution]

https://kb.netapp.com/on-prem/ontap/DP/SnapMirror/SnapMirror-KBs/How_to_move_a_snapmirror_volume_to_another_aggregate

The main recommendation is to make sure snapmirror update is NOT running while the move is happening

So make sure you quiesce snapmirror relationship

[BenBrazil]

Thanks for the replies Rvan and Sergey.

[TMACMD]

The cleanest method would be to create intercluster LIFs on every node on each side. Make sure the cluster peer flips back to available/OK.

Then replication will continue period. When you are ready to decomm the old nodes, just delete the old InterCluster LIFS. You may need to update the cluster peering partner IP lists on each side.