NetApp Data OnTap 8 comes with a feature called Data Motion which will move Volumes between aggregates with no disruption.
But for places that are running OnTap 7.x, and need to migrate Volumes from one aggr to another there is ndmpcopy or SnapMirror.
I had the task of moving all data from old NetApp shelves into new Shelves, this really meant to me migrating Volumes from the aggregates in the old shelves into aggregates in the new shelves.
For this guide I am going to use SnapMirror and the task is to migrate the volume ‘oldvol’ sitting on the aggregate ‘oldaggr’ to volume ‘newvol’ which will seat on aggregate ‘newaggr’. All of this is happening on the same NetApp controller, I am not migrating to another controller in this instance, this is just to decommision the old shelves.
Filer1:oldaggr:oldvol->Filer1:newaggr:newvol
1) Check that you have SnapMirror license
filer> license
snapmirror XXXXXX
* If you don’t you will need to purchase one and install it.
2) Add the controller (in this case is the same) to the allowed SnapMirror Hosts
options snapmirror.access host=filer1
3) Enable SnapMirror
options snapmirror.enable on
4) Create the SnapMirror destination volume. The size of the destination volume must be at least the same size as the original volume
vol create newvol newaggr 100G
// The original volume oldvol is also 100G
5) Restrict your destination volume to leave it ready for SnapMirror
vol restrict newvol
6) You can schedule replication to happen often, that way when you are ready to migrate, less data will need to be migrated during the cut-over. I ran scheduled replication every night at 10:00 PM, let it run during weekdays and cut-over to the new location on Saturday morning.
Add the schedule to /etc/snapmirror.conf
FILER1:oldvol FILER1:newvol – 0 2 * *
7) At this point we are ready to start the SnapMirror relationship
snapmirror initialize –S FILER1:oldvol FILER1:newvol
8 ) Monitor status of the replication
snapmirror status
9) At this point we are ready to cut-over to the new shelves/aggr. If you have a LUN in the volume, you might want to disconnect the server that attaches to the LUN by either disconnecting the LUN or unmapping the LUN to the server, or bring down the server while you are doing this maintenance.
10) Now run the migration, which will do the following:
- Performs a SnapMirror incremental transfer to the destination volume.
- Stops NFS and CIFS services on the entire storage system with the source volume.
- Migrates NFS file handles to the destination volume.
- Makes the source volume restricted.
- Makes the destination volume read-write.
filer1> snapmirror migrate oldvol newvol
11) snapmirror migrate will migrate NFS handles, but you will need to re-establish CIFS connections and map the igroups to the new LUN paths