The great virtualisation project mentioned in the last post hasn't really kicked off as I had envisioned and no hardware has made an appearance yet! From the outset this blog has been a roller coaster ride of exciting facts so apologies to my avid readers for the delay here.
Had a simple server move to do yesterday. We moved a switch, a storage server and its associated SAS-attached storage and an ESXi server from one data center to another. A NetApp FAS2050 was decommissioned and lies unused on our server room floor.
The decommisioned 2050 was the snapmirror target but was replaced by a hosted "vFiler". I intend on powering up the old one soon, but what do I have to do to it beforehand? Delete snapshots, that's what:
Show all snapshots on the current controller:
snap list
Show the status of the various snapmirrors:
snapmirror status
Delete a particular snapshot:
snap delete [volume_name] [snapshot name]
One of the snapmirrors to the decommisioned SAN was "busy" and so the snapshot could not be deleted. I used
snapmirror abort [source_filer]:[volume_name] [destination_filer:volume_name]
But it told me that it was already idle. So how could I delete a snapshot that was supposedly busy (but was not apparently involved in a snapmirror)?
Well it turns out that that volume was being snapmirrored to the new hosted vFiler and this was causing a lock. I'm not sure, since there are two snapshots for that volume, one for the old destination and one for the new destination, why this would happen but there you go.
No comments:
Post a Comment