It depends if you create the aggregate on the destination filer first, if you do then steps are as follows:
1) Remove the shelf from dstfiler and connect to srcfiler
2) On srcfiler run the command disk reassign -s <dstfiler sysid> -d <srcfiler sysid>
3) The destination aggregate and volume should now be visible. If they have the same name as an already existing aggregate/volume then they'll have been marked as foreign and offlined.
4) On srcfiler issue the command `snapmirror initialize -S srcvolume srcfiler:dstvolume`
5) When complete issue `snapmirror break dstvolume` Issue `snap list` on both srcvol & dstvol to confirm a common snapmirror snapshot exists.
6) Remove the shelf from srcfiler and connect to dstfiler
7) On dstfiler run the command disk reassign -s <srcfiler sysid> -d <dstfiler sysid> The destination aggregate and volume should now be visible.
😎 Add an entry into /etc/snapmirror.conf on dstfiler such as `srcfiler:srcvolume dstfiler:dstvolume restart=always 05 * * *` (For snapmirror.conf syntax see http://now.netapp.com/NOW/knowledge/docs/ontap/rel736/html/ontap/onlinebk/GUID-47724B21-7C7E-4AC9-A497-9505B42C27AF.html)
9) On dstfiler issue the command `snapmirror resync dstvolume` You should get a message stating which snapshot will be used as per the common snapshot in step 5.
10) Once complete resyncs should take place as per schedule in snapmirror.conf or in Protection Manager.
This is from memory so I may have missed a couple of minor things that will be obvious, the basics should be correct though.