Yes, the reversion did finish after 80 minutes (for four 72 GB drives in an F760), and now I can happily vol copy from the 230.
Cool!
During the migration, though, my destination volume on the F760 will be a newly created volume of ten 18 GB drives--I will be unable to create that volume until I first migrate data from the existing volumes using those drives. Therefore, I won't want to create my new volume under ONTAP 6.0 and wait for the reversion before I can continue my data migration.
Under 6.0, is there a way to create a new volume without checksums instead of waiting for a forced reversion?
Not that I'm aware of. Remember, it's not just checksums, there are other filesystems changes (like the split blockmap for snapshots) that have to be converted as well. I'm honestly not sure which parts of the reversion take the most time...I'm just not that deep into WAFL.
Was there any data on the volume that took 80 minutes? I would think that an empty volume with no snapshots should be quicker than a volume with data and snapshots. Again, haven't tried this in a lab, but it makes sense.
Hope this helps...good luck!
-- Adam Fox NetApp Professional Services, NC adamfox@netapp.com
On Wed, 21 Feb 2001, Fox, Adam wrote:
Was there any data on the volume that took 80 minutes? I would think that an empty volume with no snapshots should be quicker than a volume with data and snapshots. Again, haven't tried this in a lab, but it makes sense.
I posted a similar message a few weeks ago while playing with SnapMirror between a source F740 running 5.3.6R2 and a destination F840 running 6.0.1R1. It was a virgin F840 system, freshly zeroed drives in the snapmirror volume, and it still took over an hour to revert the filesystem. I'm guessing it has to do some operation that involves marching through every block on the volume, allocated or not.
-
Brian Tao -
Fox, Adam