VMware Cloud Community
timsaunders
Contributor
Contributor
Jump to solution

Issues during a Storage vMotion

Hi,

I've just started a Storage vMotion of a VM that's total Storage is 445GB (split across 4 vmdk files).

The migration started fine, but when it got to about 20% somebody rebooted the VM. The VM will now not come back up again but the %age migrated is still rising, however I can't seem to see any progress on the file sizes.

I would like to cancel this migration and start again but whenever I right click on the active task to cancel it, the cancel option is greyed out.

Does anybody know how I can cancel this migratrion so I can start it again (without causing any issues with the vm)?

Thanks

Reply
0 Kudos
1 Solution

Accepted Solutions
JaySMX
Hot Shot
Hot Shot
Jump to solution

What files are in each location and what are their sizes? The SVmotion process creates a snapshot on the original disk, while it copies the .vmdk(s), then reparents the snapshot and commits it on the destination. If that process was interrupted, the VM may be in an inconsistent state on the remote datastore.

Check this link as well: http://communities.vmware.com/thread/134498

-Justin

View solution in original post

Reply
0 Kudos
4 Replies
timsaunders
Contributor
Contributor
Jump to solution

Update: This has got worse.

The migration failed and I'm now not able to bring the vm back up. The vmx file and other config files have moved across to the new LUN but the vmdk files haven't (well it shows that they have, but I can't use them). The vmx file was pointing to the new LUN's for the vmdk files, so I cahed that to point back to the original LUN's, but no luck. Does anybody know how I can get this server to boot back up using it's old vmdk files on the original LUN's?

This has become urgent now

Reply
0 Kudos
JaySMX
Hot Shot
Hot Shot
Jump to solution

What files are in each location and what are their sizes? The SVmotion process creates a snapshot on the original disk, while it copies the .vmdk(s), then reparents the snapshot and commits it on the destination. If that process was interrupted, the VM may be in an inconsistent state on the remote datastore.

Check this link as well: http://communities.vmware.com/thread/134498

-Justin
Reply
0 Kudos
timsaunders
Contributor
Contributor
Jump to solution

Hi,

After a bot of "fiddling" around (following http://www.van-lieshout.com/2009/03/how-to-recover-from-failed-svmotion/), I seem to have fixed it. It seems as though the vmdk files were in a DMotion state - and as soon as I followed the steps in this guide, the VM came up perfectly - it came up on the new Storage, not back on the original (as I was expecting), and a check in the evnt logs reveal no immediate issues.

I'll be keeping the original vmdk's for now, just in case!

I'll give you the point as you gave me an answer that I think would have solved it!

Reply
0 Kudos
JaySMX
Hot Shot
Hot Shot
Jump to solution

Thanks and glad you got it working, thanks also for posting the link you found. That will help people with this problem in the future.

-Justin
Reply
0 Kudos