Forum Discussion

oliveee's avatar
oliveee
Level 3
13 years ago

Hyper-V Failover Cluser backup issue

Hello to all,

I have recently setup an Hyper-V failover cluster with 2 brand new servers configured as clustered nodes along with a CSV storage.

THe cluster is working great, the failover works fine, and I can Live Migrate Virtual Machines as much as I want.

However I am  running into a Backup issue.

My Backup software is Backup Exec 2010 R3. I have installed the backup agent on my 2 failover nodes.

In the backup job selection I can see my cluster refered as Hyper-V-Cluster1 , and underneath I see C:\ClusterStorage\Volume1 , and also Microsoft Hyper-V-HA Virtual Machines , with my Virtual Machines listed.

So far so good.

The problem is that , when selecting VMs under Hyper-V HA , the backup will fail for all the VMs stored on the node which is not the storage Owner Node.

For the owner node the VMs are backedup properly as long as I choose the following Advanced Open File option:

Microsoft Volume Shadow Copy Service => Automatic, Allow Vss to select the snapshot provider.

But for the VMs on the NON storage Owner node the job will failed with the following error:

 - AOFO: Initialization failure on: "VRTSRV::\\HyperV-Cluster1.xxx.local\Hyper-V?HA?VM\VM1". Advanced Open File Option used: Microsoft Volume Shadow Copy Service (VSS).
V-79-10000-11230 - VSS Snapshot error. Unable to get minimum quiet time for physical volumes. Reduce the activity on these volumes, or wait and run the backup when there is less activity on the volumes.
The following volumes are dependent on resource: "C:" "C:\CLUSTERSTORAGE\VOLUME1" .


I cannot back up VMs stored on the non owner Hyper-V host ( node).

I  tranfered this VM on the Owner node, and then the backup  works just fine.

So I would like to know how to backup all my VMs stored on the other node.

Besides , in the job selection, if I choose my VM from the C:\Clusterstorage\Volume 1  instead of selecting it from Hyper-V HA , it works fine too.

So I am also wondering what is the difference between choosing Hyper-V HA , and C:\Clusterstorage

I am adding a picture below, so it makes more sense.

 

Thank you for your answers !

  • Hello,

    Anybody would haved a lead on that matter? I still have the same issue, and have no idea about how to fix it.

    Thanks

     

  • C:\ClusterStorage\Volume1 is the container for CSV, whereas Microsoft Hyper-V HA Virtual Machines is the container for all VMs & their config data...Are the remote agents for windows installed on all the Hyper-V hosts ? And are you selecting the VMs individually or are you selecting the entire cluster node ? Lastly, choose the Microsoft VSS provider as the explicit provider under the AOFO setting..

  • Thanks for the answer.

    Yes the remote agents is installed on both Hyper-v nodes of the cluster , which allow me to see the Hyper-V cluster name in the bakup job selection.

    The remote agents is also installed on all the VMs I am trying to backup , using the Hyper-V HA option in Backup job selection.

    I am selecting the VMs individually underneath Hyper-V Ha, since not all the VMs need to be backed up for now.

    Regarding the snapshot provider , I was used to choosing the Microsost VSS provider before, but if I would do so for the Hyper-V HA backup, I would end up with the following error message on my backup job:
    VSS snapshot provider does not support snap technology for this volume

    That's why I switched to the Automatic Allow VSS to select the snapshot provider , in the backup job options.

    Whit this option enabled, It is working fine, but only for the VMs hosted on the cluster Owner Node.

    I tried to change the owner node from Server1 to Server 2 , just to run a test. It occurs that it does not matter whether Server 1 or Server 2 is the owner node, only the VMs on the Owner nodes will be backed up.

     

     

  • Just a quick Q...Is the other node added to the list of possible owners for the CSV ?

  • Actually I assume so, given that when I put one of 2 nodes in maintenance mode, the other takes over and becomes the new owner node on the CSV.

    However I cannot access to the list of owners by editing the properties od the CSV disk in the failover cluster GUi, nor by command line :

    Get-Cluster -Name Clustername | Get-ClusterResource | Get-ClusterOwnerNode

    The CSV disk would not show up in the list.

    But anyway I assume the 2 nodes are by definition owners , since they can take over each other, when one serve is down.

     

  • Run this command & pls check -

    Cluster res <resource name> /listowners

    Example: Cluster res CSV_1 /listowners

  • Hello,

    The problem is likely to be fixed.

    The backup seems to be working now, so I'll describe below what has worked for me:

    First of all , I realized that the 2 Remote agents installed on my 2 clusters nodes were not up to date.

    For some reason the remote installation from the Backup Exec Media Server was not working, so I had to install it locally on the 2 nodes using the Backup Exec 2010 DVD we had.

    The problem is that the Remote Agent I was installing from the DVD was not the same version as the one available on the Backup Exec Media server, which I could not deploy from the network.

    I had to manually copy the Remote Agent from the Backup Exec Media server using the following procedure:

    http://www.symantec.com/business/support/index?page=content&id=TECH158716

    And then install this version on the 2 clusters nodes. Once it was done the backup was still not working, but I had other error messages. I also had to disable the antivirus on the 2 nodes to make it run . I already did that with the old remote agents but it was still not working. I will find a way to set up the antivirus not to make the backup fail.

    Thanks

     

  • Unfortunatelly I thought it was OK, but it is not.

    I realized that instead of selecting the VMs from Hyper-V HA, the selection was pointing to C:\Clusterstorage. That is why it was working.

    Since we have a Backup Exec 2012 license, I decided to upgrade to Backup Exec 2012 to see if it makes any difference. I have also uploaded the latest Remote Agent to the 2 clusters servers.

    But I am still getting the very same error messages.

    Since we are using Windows Server 2012 we cannot use cluster.exe command line anymore.

    Instead , I used the following powershell command line:

    Get-Cluster -Name Clustername | Get-ClusterResource | Get-ClusterOwnerNode

    With the below results :

    PS C:\Windows\system32> Get-Cluster -Name HyperV-Cluster1 | Get-ClusterResource | Get-Clusterownernode

    ClusterObject                                               OwnerNodes
    -------------                                               ----------
    Cluster Disk 2                                              {Server01, Server02}
    Cluster IP Address                                          {Server01, Server02}
    Cluster Name                                                {Server01, Server02}
    Virtual Machine Configuration xxxxxxxxxxxxx                 {Server01, Server02}
    Virtual Machine Configuration xxxxxxx                       {Server01, Server02}
    Virtual Machine Configuration xxxxxxxx                      {Server01, Server02}
    Virtual Machine Configuration xxxxxxxxxxx                   {Server01, Server02}
    Virtual Machine Configuration xxxxxxx                       {Server01, Server02}
    Virtual Machine xxxxxxxxxx                                  {Server01, Server02}
    Virtual Machine xxxxxxxxx                                   {Server01, Server02}
    Virtual Machine xxxxxxxx                                    {Server01, Server02}
    Virtual Machine xxxxxxxxx                                   {Server01, Server02}
    Virtual Machine xxxxxxxx                                    {Server01, Server02}

    Server 01 and 02 are our 2 clusters nodes.

    Cluster Disk 2 resource is our Quorum disk Witness.

    Cluster Disk 1 is our Clustered Shared Volume, but it is not showing in the list.

    However it looks lie this CSV is well configured in the cluster, since the failover is working fine when we shutdown one of the 2 nodes.
    Any Feedback on that would be highly appreciated.

  • oliveee,

    I have the exact same enviroment and the exact same error. Were you able to get this resovled?

    Using BE 2010 R3, trying to backup Hyperv 2012 Clustered VMs. Remote agents are installed on the Hyperv hosts and VMs. The backups ran fine prior to creating the cluster. Once the cluster was created the VMs do not show under the individual servers. They are listed under the Cluster node name. I'm selecting the VMs under the Cluster node in the selection list as required.

    AOFO: Initialization failure on: "VRTSRV::\\HyperV-Cluster.xxx.local\Hyper-V?HA?VM\VM1".  Advanced Open File Option used: Microsoft Volume Shadow Copy Service (VSS).
    V-79-10000-11230 - VSS Snapshot error. Unable to get minimum quiet time for physical volumes. Reduce the activity on these volumes, or wait and run the backup when there is less activity on the volumes.
    The following volumes are dependent on resource: "C:" "C:\CLUSTERSTORAGE\VOLUME1" .

    I could upgrade to BE 2012, but from your post above, it didn't appear to resolve your issue.

    Let me know if there is a solution.

     

    • ashrafa's avatar
      ashrafa
      Level 1

      Dear I am facing the same issue as you have mentioned above, were you able to resolve above hyperv cluster backup issue?

      I would really apperciate if you could help me out.