2013-07-26 08:45 AM
We are about to migrate several volumes from one aggregate to another but before going ahead using the vol move command I ran a quick test on a couple of servers just to be sure. The migration worked fine and the data was accessible still post migration and in fact didn't go offline at all as both aggregates are on the same controller.
The problem comes when I then try to delete the disk from snapdrive for windows as it states 'The system cannot find the path specified' and errors out when I attempt to do this.
Before I go ahead and migrate any of our production volumes can anyone tell me if this is expected behavior in this scenario, and if so, how do I get snapdrive to re-connect to the disk at the new location post migration in order to re-size etc as required in the future?
If this is not expected...does anyone know what might be wrong?
Solved! SEE THE SOLUTION
2013-07-26 08:53 AM
Hi - Why would you want to delete the LUN through Snapdrive, if you have vol moved to another aggr?. Did the LUN go offline during the vol move at all?
Sounds like Snapdrive has some stale metadata regarding the vol/LUN location. Was SDW restarted on the guests, or the guest rebooted at all?
2013-07-26 08:55 AM
I have done this kind of migration many times, but I used snapmirror instead of vol move.
in addition I did not use snapdrive delete disk, used snapdirve disconnect disk.
in this scenario you have both drives at the same time in case something goes wrong.
E.g. disconnect from old drive and connect disk to the newly created vol lun etc...
I can post / more detailed steps if needed.
2013-07-26 09:00 AM
The moves I performed were on a test system and all went OK. I was just trying to tidy up the test server when this error occurred which made we wonder if there was going to be some issues when we perform this on live data volumes which would still need to be accessed and managed through snap drive post migration.
2013-07-26 09:04 AM
We're going to have to use snapmirror on some of the volumes which reside on an aggregate on the other controller in the HA pair and then I guess we will need to have an outage in order to connect to the new volumes. In this instance I was trying to use vol move to migrate the data 'online' with no outage to the server. This worked fine but like I said, there may be an issue with the snapdrive talking to the volume again post migration which would be a problem for us as this is used to manage disk sizes etc.
2013-07-26 09:13 AM
Thanks Martin for your input.
It seems that the snapdrive service just needed to be restarted and it picked up the new location of the disks just fine and a test expansion worked as hoped!