none
Cannot protect all Hyper-V virtual machines RRS feed

  • Question

  • Have installed System Center DPM 2012 SP1 with Rollup 1 (KB 2785682) on a 2012 Server. I'm unable to protect several Hyper-V VM's:

    Problem 1:

    Cannot protect three VM's running on Hyper-V 2012 (but eight VM's can be protected). Getting the following error: DPM is unable to continue protection for \Backup Using Child Partition Snapshot\<vm> on <hyper-v-host>.com because the change tracking information is corrupt (ID 30156 Details: Do not attach the filter to the volume at this time (0x801F000F)). 

    Tried synchronization job with consistencty check, does not work. Have removed the VM from the protection group and deleted the replica on the backupdisk. After that, have added the VM to the protection group again. Initial synchronization starts, but the same error that I mentioned above appears again.

    How can we establish protection for these VM's?

    Problem 2:

    We have an important Windows 2003 VM with SQL (it's a Microsoft CRM Server). DPM only offers protection using saved state which is not acceptable for the customer. The VM fullfills all requirements for DPM for protection using child partition, as documented on Technet, http://technet.microsoft.com/en-us/library/dd637102.aspx.

    Thank you in advance for any advice.

    Franz


    Tuesday, April 2, 2013 10:01 AM

All replies

  • Hi,

    For problem 1 - Is this a Hyper-V cluster and if so are the effected VM's all on the same CSV ?

    For problem 2 - DPM is not involved In deciding what state a VM gets backed up in.  It's the Windows Hyper-V writer that communicates with the Integration components inside the guest.  Make sure integration components are up to date in the guest, and there is at least 300MB free space on each volume inside the guest.

    To get a real-time view of how then next backup will be performed, you can run the following test.

    To list hyper-V writer captions for guests to see if online or saved state backups can be done:

    From an administrative command prompt:

    a. Type: copy con script.txt and press enter
     b. Type: list writers and press enter
     c. Press CTRL + Z (to save)
     d. Type: diskshadow /s script.txt | find /i "caption: backup using"

    Output will be displayed for each guest.

    Since the problem is outside of DPM, I recommend you start a new post in the Windows Clustering forum:
     http://social.technet.microsoft.com/Forums/en-US/winserverhyperv/threads


    Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread. Regards, Mike J. [MSFT] This posting is provided "AS IS" with no warranties, and confers no rights.

    Wednesday, April 3, 2013 9:50 PM
    Moderator
  • Hi

    Thank you for your support. I will troubleshoot problem 2 and post the results here.

    Details for problem 1: It's not a cluster, therefore no CSV's. We have two Server 2012 Hyper-V hosts (and one 2008 R2 SP1 Hyper-V Host), all with local storage and enabled replication for all important VM's.

    Franz


    Thursday, April 4, 2013 6:58 AM
  • Franz, have you checked the Hyper-V host eventlogs? Specifically the Hyper-V VMMS eventlogs. Very often you can find clues in there why a backup of a virtual machine doesn't work. Also check the SQL server machines eventlogs.
    Thursday, April 4, 2013 9:07 AM
  • Hello and thank you all for your support!

    - Problem 1: (unable to protect VM's on one Hyper-V 2012 Host): I could restart this Hyper-V host over lunch time. Have examined all Hyper-V eventlogs. There is no warning or error eventlog entry when trying to run a resynchronization job in DPM. Just information events that the VSS service was started. Hyper-V 2012 runs solid, there aren't any warnings or errors concerning Hyper-V on this server.

    - Problem 2: (Only "save state" protection offered by DPM): Mike was right, it seems to be a Hyper-V problem on this 2008 R2 hyper-V Host. Executing diskshadow, then "list writers" provides the following output:

                    + Component "Microsoft Hyper-V VSS Writer:\881FBAAA-0AB4-4CF5-BA
    8C-69BDE837B709"
                            - Name: 881FBAAA-0AB4-4CF5-BA8C-69BDE837B709
                            - Logical path:
                            - Full path: \881FBAAA-0AB4-4CF5-BA8C-69BDE837B709
                            - Caption: Backup Using Saved State\cm9
                            - Type: VSS_CT_FILEGROUP [2]
                            - Is selectable: TRUE
                            - Is top level: TRUE
                            - Notify on backup complete: FALSE
                            - Paths affected by this component:
                                    - d:\Virtual Servers\cm9\
                                    - d:\Virtual Servers\cm9\Virtual Machines\
                                    - d:\Virtual Servers\cm9\Virtual Machines\881FBA
    AA-0AB4-4CF5-BA8C-69BDE837B709

    Have no idea why Hyper-V is offering only save state protection for this VM.

    Franz

    Thursday, April 4, 2013 12:23 PM
  • Hi Franz,

    For problem-2 - most likely it's an issue inside guest itself since other VM's are not effected.  Open a new question for that issue here:  http://social.technet.microsoft.com/Forums/en-US/winserverhyperv/threads

    For problem-1 - it sounds like the dpmfilter has detected corruption in the filter bitmaps.  You can try this to force the dpmfilter to make new ones.

    Open an administrative command prompt on the Hyper-v Server.

    1- Stop the DPMRA:
        Net stop dpmra

    2- Unload the dpmfilter:
        Fltmc unload dpmfilter

    3- Set the Force Fixup Registry key
         Reg add "HKLM\SOFTWARE\Microsoft\Microsoft Data Protection Manager\Agent\2.0" /v ForceFixup /t REG_DWORD /d 1

    4- Run a consistency check for the effected VM's to get them consistent.

    5- Let some normal recovery points run - they may take longer to complete - but they should succeed.  Wait for 1 day to get good recovery points.

    6- The next day - Stop DPMRA again then Load the dpmfilter at an elevated command prompt using:
        fltmc load dpmfilter

    7- Remove the Forcefixup key from the registry.

    8- Run a manual CC on all the hyper-v VM's again, that should start with new bitmaps.

    9- See if normal recovery points work.


    Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread. Regards, Mike J. [MSFT] This posting is provided "AS IS" with no warranties, and confers no rights.

    Thursday, April 4, 2013 3:43 PM
    Moderator
  • Hi Mike

    Thank you!

    Have executed step 1 to 4, and could manually successfully create recovery points for the VM's the first time.

    But I will be abroad for the next two months. Have forwarded the information to our customer, but it is not certain that the customer will execute the steps 5 to 9. Our customer seems to prefer to pause the DPM project until my return in mid June, in the hope that DPM 2012 SP1 gets "more mature" and that some of our problems will be solved with less effort (and that all the Weblinks in DPM start to work and do not lead either to the page with evaluation Software, or to "page not found" when clicking the hyperlink in a DPM message).

    Kind regards,

    Franz

    Friday, April 5, 2013 5:03 PM
  • Hi Mike

    Franz is at the moment in holidays and i will look at this time for this problem. We have make your steps. Now the replication of these 3 servers are inconsistent i there is following error:
    An unexpected error occurred while the job was running. (ID 104 Details: Do not attach the filter to the volume at this time (0x801F000F))

    While we tested your steps, we had 1 virtual server, who used snapshots and he had massive problems with the network after going back to a snapshot. it was impossible to grant him access to the internet or internal ressources. After your steps 5 - 9 and restarting the Hyper-V Server, the network is working normal with this machine.

    Do you have an idea, why the replication is inconsistent and how we could solve this problem?

    Kind regards,
    Manuel Kobel

    Thursday, April 11, 2013 1:39 PM
  • Hi,

    So what I think you are saying is if you just follow steps 1-4 - then backups are successful, but as soon as you re-introduce the DPMFilter driver, the backups fail once again.

    We just introduced UR2 for DPM 2012 SP1, please install that and see if that helps in any way, if not, please open a support case for additional troubleshooting.

    System Center 2012 SP1: Update Rollup 2 for System Center 2012 Service Pack 1
    http://support.microsoft.com/kb/2802159


    Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread. Regards, Mike J. [MSFT] This posting is provided "AS IS" with no warranties, and confers no rights.

    Thursday, April 11, 2013 3:40 PM
    Moderator