locked
Drive dissapears after Hyper-V Backup RRS feed

  • Question

  • Hello all,

    This is actually part of a larger issue that we discovered recently when users tried to login but their profile couldn't load nor could they access their redirected folders. At that time we discovered that a drive on our Domain Controller had vanished. Upon investigation we discovered within the Hyper-V settings that the disks currently attached were differencing disks which were created automatically when the backup ran.

    To resolve this issue at that time we initiated a shutdown of the VM which triggered a merge. This caused the original avhdx file to be attached again and once booted the drive/volume was als available again within the VM environment itself.

    So my question is twofold basically:

    1. What could cause this drive to suddenly dissapear within the VM itself after the backup runs?

    2. Why is Windows Server Backup not merging the checkpoints after the backup has finished? It doesn't seem to do this for any of the VM's which means we manually have to initiate a shutdown to trigger the merge. Considering that we run a daily backup having to do this is far from desirable.

    On a sidenote, after the last backup everything still seems to be working but I suspect the problem of the dissappearing drive only occurs when the second backup occurs. To be more clear:

    1. Original ahvdx files are attached to Virtual Machine(s)
    2. Windows Server Backup runs and creates a checkpoint
    3. Backup finishes but merge is not initiated
    4. Virtual Machine is NOT shutdown manually
    5. Windows Server Backup runs again, this time with the differencing disk attached to the VM
    6. Drive dissapears from within the Virtual Machine

    Note that at this time this is only the above is only an assumption. We will know for sure tomorrow after the backup has run again. But the problem still remains that sometimes, after a backup, our data drive dissapears from within the VM and that Windows Server Backup does not automatically perform a merge after it has completed. Now these two issues might not actually be related but it would be a really strange coincidence.

    We are running Server 2012R2 Standard with two Virtual Machines. We use Windows Server Backup to perform a daily backup on the Hyper-V host with the following items included:

    • Bare metal recovery
    • System state
    • Host component
    • Virtual Machine 1
    • Virtual Machine 2
    • EFI System Partition
    • Local Disk C
    • Recovery

    Oh and perhaps useful to know, we do not see any errors on the Hyper-V host regarding the backup. We do see always see event id 157 - Disk x has been surprise removed when the backup runs.

    Wednesday, January 11, 2017 10:44 AM

Answers

  • Actually the problem seems to have disappeared. One of the Virtual Machines had already started working right with checkpoints being merged after a backup. The other still had a lingering checkpoint from a week or so ago. So I double checked to make sure it was safe to delete while the VM was running, deleted it, and ran another manual backup. Checkpoints were created for both Virtual Machines, the backup was run (and finished), and the checkpoints were succesfully merged into the existing VHDX files.

    So I have no clue why it's been resolved suddenly (we haven't installed any updates yet) but sometimes you just gotta accept the magic I guess!

    Thursday, January 26, 2017 7:59 AM

All replies

  • Hi Roy,

    The issue seems strange.

    Which OS is the VM running?

    Have you install integration services?

    If you create a checkpoint and delete manually, is it merging automatically?

    Best Regards,

    Leo


    Please remember to mark the replies as answers if they help.
    If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.

    Thursday, January 12, 2017 8:34 AM
  • Hello Leo,

    My apologies for the lack of a reply but we're in the middle of several projects so this had kind of slipped my mind. I will test the manual checkpoint somewhere tonight (GMT+1 timezone) or early this weekend and let you know my findings. It might be safe to do it during the day but I'd rather not take the chance so I'm waiting until everyone there is done for the day so I can safely test without causing any issues.

    Friday, January 20, 2017 10:31 AM
  • Hi Roy,

    Are there any updates?

    Best Regards,

    Leo


    Please remember to mark the replies as answers if they help.
    If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.

    Monday, January 23, 2017 2:18 AM
  • Hi Roy,

    Do you have mcAfee installed on your PC?

    I have the same problem.

    David

    Wednesday, January 25, 2017 4:17 PM
  • Actually the problem seems to have disappeared. One of the Virtual Machines had already started working right with checkpoints being merged after a backup. The other still had a lingering checkpoint from a week or so ago. So I double checked to make sure it was safe to delete while the VM was running, deleted it, and ran another manual backup. Checkpoints were created for both Virtual Machines, the backup was run (and finished), and the checkpoints were succesfully merged into the existing VHDX files.

    So I have no clue why it's been resolved suddenly (we haven't installed any updates yet) but sometimes you just gotta accept the magic I guess!

    Thursday, January 26, 2017 7:59 AM