VMware SRM–After failover, LUNs still have ‘snap-xxxxxxx’ in the name
VMware’s Site Recovery Manager is a great product! Allowing companies to protect and failover in the event of a disaster is amazing. The ability to test these plans in a bubble is what really sets it apart for me though. For my clients, we do bubble tests quite often to make sure our runbooks are good and we are protecting everything we need in the event of a disaster. Even with all this planning, every so often we need to run a full failover to make sure our planning and testing are complete and valid.
A noticable difference in bubble testing and full failover is the preservation of LUNs during a full failover. When we do bubble tests, once the testing and validations are complete, we toss the LUNs in the DR and are ready to go again in Production. For Full Failovers though, we must go through the reprotect process and actually do the failback.
At a client, I ran into an interesting issue where during the failback (and failover), the LUNs that were failed over were prefixed with snap-xxxxx-[DATASTORE-NAME]. No other functional difference in the LUN or virtual machines but not what the admins expect to see in the vCenter screen. If the failback is successful, you can just rename the LUN back to it’s original name. No downtime or mess.
If you would like SRM to rename the LUN to avoid this in the future, there are 2 relevant settings in the advance section :
Sites –> [Primary Site] –> Manage –> Advanced Settings –> Storage Provider –> EDIT
From there, you can find the following section and modify appropriately
Once this setting is set to enabled, SRM will automatically rename the LUNs after a successful failover.