locked
Incomplete VM Configuration RRS feed

  • Question

  • Hi All,

    We have an implementation of Hyper-V on Server 2012 R2 and is being managed by SCVMM 2012 R2. Storage is presented to Hyper-V hosts through a Scale-out file server cluster. The issue is, every now and then few VMs go into Incomplete VM Configuration status. VM continue to run perfectly and status is ok in Hyper-V manager and Failover cluster manager. If I try to run the Repair option in SCVMM, the resulting windows shows up empty and repair button greyed out.

    Done some more digging. On Hyper-V host, I see following events getting logged.

    Log: Microsoft-Windows-Hyper-V-Worker/Admin
    Event ID: 12292
    Message:

    'VM01' encountered an error while calculating the size of the disk chain for '\\FS-Cluster\Production\VM01\WIN2012R2DataCenter.vhdx'. The error was: 'The process cannot access the file because it is being used by another process.' (0x80070020). Maximum disk metric will not be updated for the current disk. (Virtual machine ID A6F542DE-D201-4F2F-A11C-C2422C5C05B5)

    Log: Microsoft-Windows-Hyper-V-VMMS/Admin
    Event ID: 15268
    Message:

    Failed to get the disk information.

    Log: Microsoft-Windows-Hyper-V-VMMS/Storage
    Event ID: 27000
    Message:

    Failed to open attachment '\\FS-Cluster\Production\VM01\WIN2012R2DataCenter.vhdx'. Error: 'The process cannot access the file because it is being used by another process.'.

    I already have made sure that

    - No antivirus software is present on Hyper-V host or on scale out file servers
    - The Integration services are upto date on these VMs

    What I do to resolve this is live migrate virtual machine using failover cluster manager and refresh the VM in SCVMM. It then updates the status back to running but getting annoyed of having to do it again and again for no apparent reason. Any ideas ?

    Thursday, March 10, 2016 8:34 PM

Answers

  • Are you making changes to the VMs outside of VMM (FCM \ Hyper-V Manager etc...) If so VMM can get a little stroppy. When we have seen this in the past we have performed the following:

    Refresh host cluster (In VMM)
    Refresh virtual Machines (On each cluster node in VMM)
    Right click the VM > Repair > Ignore

    This normally clears it.

    Sounds to me like VMs are being migrated \ changed outside of VMM and your VMM is not using events based notifications - ie. updates to VMs inside of VMM are occurring on a schedule rather than by event notifications.

    this should tell you if any of your nodes are running in "Legacy" "timer based" mode.

    Get-SCVMHost | Where-Object -FilterScript {$_.GetRefresherMode() -eq "Legacy"}

    What is this VHDX  '\\FS-Cluster\Production\VM01\WIN2012R2DataCenter.vhdx'?
    Is it a template? the OS disk of a VM?

    What update rollup of VMM are you running? In the early days there were issues with this, UR8 and UR9 are *much* better.

    Might also be worth looking at this:

    https://support.microsoft.com/en-us/kb/3001854 (how to enable events based refresher in VMM)

    Important Note:

    Please be aware that, after you install System Center 2012 R2 Update Rollup 6 (SC 2012 R2 UR6), you must make an additional change in order to enable the correct refreshing of virtual machines across hosts. If you are using System Center 2012 R2 Update Rollup 6 or a later version, you must add the following registry values on the Virtual Machine Manager server:
    Registry location: HKEY_LOCAL_MACHINE\Software\Microsoft\Microsoft System Center Virtual Machine Manager Server\Settings
    Registry key: VMPropertiesEventAssitedUpdateInterval
    Registry type: DWORD
    Min value: 0 seconds
    Max Value: 20 days

    If this is NOT set to 120 or is missing, then the following powershell will add it. (It has been known to be set to 20 days, don’t ask how many seconds that is…)

    New-ItemProperty 'HKLM:\software\microsoft\Microsoft System Center Virtual Machine Manager Server\Settings' -Name VMPropertiesEventAssitedUpdateInterval -Value 120 -PropertyType "DWord" -Force


    Yes, there’s a typo in the key!!!


    This posting is provided "AS IS" with no warranties, and confers no rights. Please remember, if you see a post that helped you please click "Vote as Helpful", and if it answered your question, please click "Mark as Answer". I do not work for Microsoft, I manage a large estate in the private sector, my views are generally first hand production experiences.



    • Proposed as answer by Microsoft Jim Monday, March 21, 2016 2:36 PM
    • Edited by Microsoft Jim Monday, March 21, 2016 2:38 PM
    • Marked as answer by Elton_Ji Sunday, March 27, 2016 2:10 PM
    Monday, March 21, 2016 2:35 PM

All replies

  • Hi Sir,

    When the issue arises , please check the storage status for that cluster node :

    http://whpwtfdidn.blogspot.jp/2013/11/incomplete-vm-configuration-virtual.html

    In addition , please configure "power management " to "not save power" for  physical NICs of physical nodes .

    Also try to disable VMQ for troubleshooting .

    Best Regards,

    Elton


    Please remember to mark the replies as answers if they help and unmark them if they provide no help. If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com .

    Monday, March 14, 2016 10:25 AM
  • Hi Elton,

    Thanks for the suggestions. I tried the blog post and but I don't have any storage and network issue on host. Other VMs are running perfectly on the same host. Power management is disabled for NICs and we already have VMQ disabled (to troubleshoot another problem). Still scratching my head.

    Tuesday, March 15, 2016 3:02 PM
  • Are you making changes to the VMs outside of VMM (FCM \ Hyper-V Manager etc...) If so VMM can get a little stroppy. When we have seen this in the past we have performed the following:

    Refresh host cluster (In VMM)
    Refresh virtual Machines (On each cluster node in VMM)
    Right click the VM > Repair > Ignore

    This normally clears it.

    Sounds to me like VMs are being migrated \ changed outside of VMM and your VMM is not using events based notifications - ie. updates to VMs inside of VMM are occurring on a schedule rather than by event notifications.

    this should tell you if any of your nodes are running in "Legacy" "timer based" mode.

    Get-SCVMHost | Where-Object -FilterScript {$_.GetRefresherMode() -eq "Legacy"}

    What is this VHDX  '\\FS-Cluster\Production\VM01\WIN2012R2DataCenter.vhdx'?
    Is it a template? the OS disk of a VM?

    What update rollup of VMM are you running? In the early days there were issues with this, UR8 and UR9 are *much* better.

    Might also be worth looking at this:

    https://support.microsoft.com/en-us/kb/3001854 (how to enable events based refresher in VMM)

    Important Note:

    Please be aware that, after you install System Center 2012 R2 Update Rollup 6 (SC 2012 R2 UR6), you must make an additional change in order to enable the correct refreshing of virtual machines across hosts. If you are using System Center 2012 R2 Update Rollup 6 or a later version, you must add the following registry values on the Virtual Machine Manager server:
    Registry location: HKEY_LOCAL_MACHINE\Software\Microsoft\Microsoft System Center Virtual Machine Manager Server\Settings
    Registry key: VMPropertiesEventAssitedUpdateInterval
    Registry type: DWORD
    Min value: 0 seconds
    Max Value: 20 days

    If this is NOT set to 120 or is missing, then the following powershell will add it. (It has been known to be set to 20 days, don’t ask how many seconds that is…)

    New-ItemProperty 'HKLM:\software\microsoft\Microsoft System Center Virtual Machine Manager Server\Settings' -Name VMPropertiesEventAssitedUpdateInterval -Value 120 -PropertyType "DWord" -Force


    Yes, there’s a typo in the key!!!


    This posting is provided "AS IS" with no warranties, and confers no rights. Please remember, if you see a post that helped you please click "Vote as Helpful", and if it answered your question, please click "Mark as Answer". I do not work for Microsoft, I manage a large estate in the private sector, my views are generally first hand production experiences.



    • Proposed as answer by Microsoft Jim Monday, March 21, 2016 2:36 PM
    • Edited by Microsoft Jim Monday, March 21, 2016 2:38 PM
    • Marked as answer by Elton_Ji Sunday, March 27, 2016 2:10 PM
    Monday, March 21, 2016 2:35 PM