So I broke a replication mirror, snap vault, etc..When I check snapmirror show the replication pair is gone.
I then took the destination volume and rehosted to a new SVM. Now normally the volume would be "RW" but for whatever reason this time on the new SVM it is not. Is there absolutely no way to convert a DP volume to RW?
If not, then I take it I can rehost the volume back to it's original destination SVM and initialize again. Not sure like I said, why the volume is nor automatically RW.
This is not possible for mirror type 'Vault' Destination.
The only time it is possible to change a -type DP volume into a -type RW (read-write) volume, is if the volume is a DP SnapMirror Destination volume and the mirror relationship is broken. An existing DP volume, which is a SnapMirror Destination volume, automatically becomes a RW volume and then you use it with existing SVM or re-host with different SVM it will remain RW, however this does not apply to mirror type 'vault'.
Note: Further, one cannot use the original destination volume as the target for the new baseline transfer. This restriction has been placed for safety reasons. Clustered Data ONTAP SnapMirror does not allow baseline transfers to a volume containing data to prevent data loss.
I guess, the workaround would be: 1) Go to your snapvault destination volume in the GUI (As long as you have flexclone licensed, you will see an option for clone) 2) Select the volume | from top menu select 'clone' and then clone the volume (Cloned vol automatically becomes RW - and so mountable) 3) Split the clone, so that the flexcloned volume becomes independent RW vol. 4) Delete the original vault dest volume (Which was DP type) 5) Use the new splitted independent volume with existing SVM or re-host with any SVM you wish to have as RW.
Ok, so I figured out what you meant by "split" the clone. I just have a hard time believing that people don't regularly have a need to create a clone from a long-term vaulted snapshot. IS there not an easier way to do that and also put that new clone on a new SVM other than the snapvault SVM?
If you need to quickly do a recover in that manner, splitting the clone can take a day if large enough or longer. I can't imagine I would need to wait that long to simply use something for which the snapvault was meant to be used.
So, in short, how can I "quickly" take a clone from a long-term (let's say 30 day) snapshot, and try to recover a file for use of recovering a VM within ESXI?