Details
-
Bug
-
Resolution: Done
-
Critical
-
None
-
7.0
-
None
-
Occurred on HA cluster of 2x XenServer 7 boxes, with 2x storage repos. One iSCSI (source), one NFS (destination)
Description
Hello,
I was taken back today when I lost all VHD's relating to a virtual machine when cancelling a VHD move from a source storage repository to a destination storage repository.
This was the scenario:
1) Virtual machine with 5x VHD's;
a) First (OS) (2tb)
b) Secondary (750gb)
c) Third (100gb)
d) Fourth (2tb)
e) Fifth (2tb)
I had moved VHD b/c/d/e to the new storage repository using the 'move' command from XenCenter, under the storage tab of the Virtual Machine in question. These moved over successfully.
I was then left with the (a) OS storage unit to move. This failed twice, and on the third go I went to the 'Notifications' tab and 'Cancelled' the move.
I then noticed (viewing the storage tab) that all the related VHD's already moved to the destination were deleted!
There must be a bug here, whereby if you cancel a VHD move, any already moved VHD's relating to the UUID of that virtual machine in question get deleted too.
I suspect the bug is: If a user cancelled a VHD move, then ensure only the VHD that is being moved is removed - not all related VHD's already at the destination related to the UUID of the virtual machine.
Sadly we lost all the data (we have backups) – but really this needs to be fixed, before more users start losing data unnecessarily.
Thanks. I hope this makes sense? This is my first post here.
Kind Regards