Last week I setup a test lun on my production NS20 and replicated it via Celerra replicator to our DR site in Wellington for the purpose of testing Vmware’s “Site Recovery Manager” which ill cover much more of in up and coming posts.
After performing fail over “tests” within SRM i finally thought right, lets put a VM on that protected lun and perform a proper failover. I hit the “Run” button agreed to the warning and watched SRM work its magic. At this point I had not figured out how to get replication working in the other direction and how to fail back so I thought “Ill just delete everything and start again”.
After shutting down the protected VM i deleted the SRM configuration and deleting the lun at the recovery site, i then went to delete the lun at the other end and found Celerra manager gave me a error of “can not delete as LUN is in use” which is something you normally see if the lun is A. either still in use… or B. The lun is still masked.. After checking this was not the case I looked at the properties of the lun and saw “Has Snapshots” = Yes so i knew at this point thats why i could not delete the lun.
I ended up using the following command to identify the full lun name and also the snapshot which i needed to delete.
(list my Luns)
# nbs-snap -a listplu -s server_2
(list the snapshot)
# nbs-snap -a listsnap -i fs36_T3_LUN0_CK200082400875_0000 -s server_2
(Delete the snapshot)
# server_iscsi server_2 -snap -delete fs36_T3_LUN0_CK200082400875__0000.ckpt001
This command completed with the usual “Done:” and then i was able to delete the LUN.