[Hyper-V] ubuntu hyper v vm freezes with high load after windows backup

[Emmanuel Günther]

Since the upgrade of the LIS tools to 4.4.0-67+ the hyper v vm backup is not possible anymore with the windows backup tool on an Ubuntu 16.04 on a Gen 2 Hyper V VM. I cannot say what is exactly happening because there are no log files and just a blank black screen. A second hyper v vm is not crashing but cannot be backuped too. Same version of Ubuntu and LIS but Gen 1 Hyper V VM.

[Brad Figg (brad-figg)]

This bug is missing log files that will aid in diagnosing the problem. From a terminal window please run:

apport-collect 1677938

and then change the status of the bug to 'Confirmed'.

If, due to the nature of the issue you have encountered, you are unable to run this command, please add a comment stating that fact and change the bug status to 'Confirmed'.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.

[Emmanuel Günther (emmanuel365)]

Right now I'm unable to run the command because I have now a backup of the vm running that backups are still possible because it's a productive system.

[Joseph Salisbury (jsalisbury)]

Are you able to perform a backup if you boot into the 4.4.0-66 kernel instead of 4.4.0-67?

[Joseph Salisbury (jsalisbury)]

This may be a duplicate of bug 1679898

[Joshua R. Poulson (jrp)]

I think this issue is with this set of storvsc changes in .67:

  * [Hyper-V] Bug fixes for storvsc (tagged queuing, error conditions)
    (LP: #1663687)
    - scsi: storvsc: Enable tracking of queue depth
    - scsi: storvsc: Remove the restriction on max segment size
    - scsi: storvsc: Enable multi-queue support
    - scsi: storvsc: use tagged SRB requests if supported by the device
    - scsi: storvsc: properly handle SRB_ERROR when sense message is present
    - scsi: storvsc: properly set residual data length on errors

We know of a regression with the last patch in this list and it is fixed with this upstream commit:

commit f1c635b439a5c01776fe3a25b1e2dc546ea82e6f
Author: Stephen Hemminger <email address hidden>
Date: Tue Mar 7 09:15:53 2017 -0800

    scsi: storvsc: Workaround for virtual DVD SCSI version

    Hyper-V host emulation of SCSI for virtual DVD device reports SCSI
    version 0 (UNKNOWN) but is still capable of supporting REPORTLUN.

    Without this patch, a GEN2 Linux guest on Hyper-V will not boot 4.11
    successfully with virtual DVD ROM device. What happens is that the SCSI
    scan process falls back to doing sequential probing by INQUIRY. But the
    storvsc driver has a previous workaround that masks/blocks all errors
    reports from INQUIRY (or MODE_SENSE) commands. This workaround causes
    the scan to then populate a full set of bogus LUN's on the target and
    then sends kernel spinning off into a death spiral doing block reads on
    the non-existent LUNs.

    By setting the correct blacklist flags, the target with the DVD device
    is scanned with REPORTLUN and that works correctly.

    Patch needs to go in current 4.11, it is safe but not necessary in older
    kernels.

    Signed-off-by: Stephen Hemminger <email address hidden>
    Reviewed-by: K. Y. Srinivasan <email address hidden>
    Reviewed-by: Christoph Hellwig <email address hidden>
    Signed-off-by: Martin K. Petersen <email address hidden>

[Joseph Salisbury (jsalisbury)]

I built a test kernel from the Xenial master-next branch, which includes the following commit:

f1c635b439a5c01776fe3a25b1e2dc546ea82e6f
Author: Stephen Hemminger <email address hidden>
Date: Tue Mar 7 09:15:53 2017 -0800

    scsi: storvsc: Workaround for virtual DVD SCSI version

The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1679898/

Can you test this kernel and see if it resolves this bug?