none
Hyper-V guest unresponsive while fixed vhdx gets expanded on iSCSI array

    Question

  • Hyper-V 2012 R2 cluster, Dell MD3600i storage array, all iSCSI connections redundant via dual 10Gb NICs & MPIO in servers & in array.

    vhdx was removed from VM & added, but before apply it was set to expand

    Resize this fixed vhdx by ie 250Gb (all that data gets written to the array) host is not affected (still can red/write to the array directly), but guest that this disk belongs to becomes completely unresponsive.

    Any idea how to tackle this slowness?

    Obviosly one is to make expansion out of hours.

    Seb

    Wednesday, May 3, 2017 11:23 AM

All replies

  • Please provide details of steps used to expand disk.  This process is generally a very quick process.

    tim

    Wednesday, May 3, 2017 12:27 PM
  • Expansion of fixed vhdx (by 250Gb) is NOT a quick process!

    Failover Cluster Manager GUI used for Expand

    Wednesday, May 3, 2017 2:17 PM
  • An offline resize is always faster.
    Wednesday, May 3, 2017 2:49 PM
  • It WAS offline resize!
    Wednesday, May 3, 2017 2:56 PM
  • "guest that this disk belongs to becomes completely unresponsive" and "It WAS offline resize!"

    Well if it was an offline resize, then of course the guest would become unresponsive.  Or do you mean that the VM has become unresponsive now that it is using the resized disk?  Not clear on what you are trying to say.


    tim

    Wednesday, May 3, 2017 10:08 PM
  • Hi Sir,

    >>Resize this fixed vhdx by ie 250Gb (all that data gets written to the array) host is not affected (still can red/write to the array directly), but guest that this disk belongs to becomes completely unresponsive.

    Do you mean the HA VM became unresponsive after you resized that VHDX file which resided in CSV ?

    Best Regards,

    Elton


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

    Thursday, May 4, 2017 9:15 AM
    Moderator
  • N, it was unresponsive DURING the process
    Saturday, May 6, 2017 7:17 AM
  • N, it was unresponsive DURING the process

    But you have mentioned it was offline resize..


    Wednesday, May 10, 2017 1:02 PM
  • I mentioned the same thing in an earlier response and scerazy did not respond to that.

    tim

    Wednesday, May 10, 2017 3:00 PM
  • Not that difficult to understand?

    vhdx was removed from VM, Apply & re-added, but before Apply again it was set to expand

    So it behaves as OFFLINE resize, it is NOT yet active for that VM (as it was not applied to the VM config)


    • Edited by scerazy Sunday, June 11, 2017 7:16 AM
    Sunday, June 11, 2017 7:16 AM
  • We are confused because you talk about the guest being unresponsive during the expansion.  And you are saying the VHDX is not active for the VM.  That's what is hard to understand.  If the VM is not attached to the VHDX, how can the VM become unresponsive?  It seems like the VM would not even be online if you have not attached the VHDX to the VM.

    tim

    Sunday, June 11, 2017 3:32 PM
  • This VM has 8 shared vhdx, so one not being attached is not really an issue for the VM to exists & be happily running!

    What else can I say? The reise made the VM unresponsive for the duration of the resize (over an hour to resize thick from 500 to 750Gb)

    Tuesday, June 13, 2017 3:16 PM