2014-12-03 10:20 AM
I've been looking at putting all of our test/dev SQL databases on to a single NFS datastore instead of what seems like an endless number of volumes and LUNs.
One thing that the SQL DBA often does is clone a DB using flexclone and attaches it to another SQL server, or the same SQL server.
What I would need to do to make this work if all the SQL DBs were VMDKs attached to a VM is to actually do a flexclone of the individual VMDK.
Is that possible? I'm not sure how I would proceed to do that. I'm OK if it's a powershell script or whatever. I just can't have a full copy operation.
2014-12-03 12:50 PM
You can make a flexclone of a file - I'm not sure if that can be done through the VSC tool or if it would need to be done via command line. There might need to be some extra coordination to ensure the activitiy on that vmdk disk is quiesced while the clone is taken.
2014-12-03 04:00 PM
Right now we're in transition, going to CDOT.
The current process is using SMSQL on 7-mode, but we've got individual volumes and LUNs for Data, logs, etc.
I don't mind doing CLI / powershell scripts. I think something like this might lead to where I want to be... (?) http://www.netapp.com/us/system/pdf-reader.aspx?m=tr-3742.pdf&cc=us
2014-12-03 04:12 PM
Yes, that TR should have the commands you'll need. At least from the point of view of generating the command to create clone of the vmdk file. The parts outside of that - of making sure the clone is good and then connecting it to another server will require a bit of creativity on your part. Not impossible to do with some help from the Vmware PowerCLI command set.