We are going to replace 2xFAS8200 with 2x60 in a DR destination cluster. NAS Volumes in FAS8200 mostly work as destination volumes in SnapMirror. There are also a handful production NFS datastores volumes.
My questions is, should using "volume move" to the new C60 be the way to migrate all volumes over?
We are thinking of using vmotion those NFS datastores over, but, since there are not too many, so, we don't see significant differences.
#Migrate all volumes from FAS8200 to C60
thick dock
uneven barn
So you currently have a 4-node FAS8200 cluster connected with cluster-switches,right? If yes, I would absolutely create an 8-node and vol move everything over, will be the easiest especially for NAS volumes. Also when it's the same cluster you don't need modify the snapmirror-relationships (only add more IC-LIFs to the cluster peering).
Storage vMotion for datastore is also fine but vol move is supperior imo since you can reuse the LIFs and volumes and don't need to create new vols, attach them to ESXi hosts, etc.
Just check via HWU if everything works, you need to stay on 9.16.1 during the migration since FAS8200 don't support 9.17.1 though. (Mixed version clusters are also supported but I wouldn't do it unless necessary.)
Also check if both clusters support the same cluster switches.
digital prawn
I try to always use storage vMotion for datastore connections. You give ONTAP, especially the c60 the opportunity to possibly improve efficiencies with little effort
uneven barn
I agree, unless you're migrating to the newer models which (at least according to documentation) will automatically do all the necessary efficiency optimizations after vol move.
https://kb.netapp.com/on-prem/ontap/DP/SnapMirror/SnapMirror-KBs/How_do_platforms_supporting_QAT_or_zlib_compression_impact_SnapMirror_and_volume_move
So with your C60 it should not be necessary anymore to do some background efficiency scans (for old data) since it's automated.