snapmirror / minimalize downtime

[TimJMcCuen]

I am looking to snap mirror volumes form an older 4 node to  a new 4 node cluster.  I plan to snapmirror volumes from one cluster to the other.    I would also like to limit downtime for my NFS hosts.     Once the data has been snapmirrored to the new cluster can reassign the data lif IPs from the old cluster to the cluster.  The new cluster will be on the same VLAN and subnet as the old cluster.    I realize at some point my hosts will probably drop the NFS connection to the NetApp storage as we transistion from old to new.    I want to minimalize downtime and not have the hosts remount the storage (data LIF) if possible.

Can this be done?  Any tips of tricks please pass my way.   Thanks.

[NetApp_SR]

Are you able to add the new nodes to the existing cluster and then use vol move?

Decide whether to upgrade by moving volumes or storage

https://docs.netapp.com/us-en/ontap-systems-upgrade/upgrade/upgrade-decide-to-use-this-guide.html

[TimJMcCuen]

Thank you for the response.   We don't want to add the new nodes to the existing cluster for warranty reasons.    We still plan to keep the old cluster cluster around til the warranty expires in a couple years.   We would like to move just a couple of volumes form the old cluster to the new with minimal disruption.   Is snapmirror the only real way to do that?   Can the Data LIFs IPs be transferred to the new cluster ?  Thanks.

[Ontapforrum · Accepted solution]

Snapmirror is ever reliable and provides data-consistent migration. In this situation, I think there will be some downtime during final cut-off (snapmirror-breaking-making dest vol read/write etc). However, NFS clients need not remount, as long as the SAME IP addresses are transferred to the New Cluster independently after the Migration.

[TimJMcCuen]

Thank you