none
Clustered disk only showing on one node RRS feed

  • Question

  • I have setup a cluster using hyper v 2008R2 It passed the all validation tests but the Clustered disks both the Quorum and one i intend for the virtual machines only load one node/server at a time. The one which boots last. If i reboot one it will load the disks and remove them from the other node. On the server withou the disk they can be seen in Disk management but are not online

    i get this error
    id 1034
    Cluster physical disk resource 'Cluster Disk 1' cannot be brought online because the associated disk could not be found. The expected signature of the disk was '{e6c8ddd8-70de-479a-8e5a-b91a4a9910b5}'. If the disk was replaced or restored, in the Failover Cluster Management snap-in, you can use the Repair function (in the properties sheet for the disk) to repair the new or restored disk. If the disk will not be replaced, delete the associated disk resource.
    Tuesday, February 16, 2010 4:35 PM

Answers

  • I would assume both nodes should be able to access the disk at the same time. 

    No, that would be the wrong assumption.

    Normally clustered disks are only accessible by the node where they are online in the cluster. The other node(s) will not have access to the disk.

    Cluster Shared Volumes (in 2008-R2) are Mounted NTFS volumes on clustered disks, but ALL nodes have access to these volumes simultaneously.

    Hope this answers your question,

    Rgds,

    edwin.

     

    Wednesday, May 19, 2010 11:59 AM
    Moderator
  • My question is, 1 )whether its normal to have this type of situation.  2) Incase of Cluster group failure on CNDL009A, and when cluster group moves to CNDL009B whether IT would be able to stay there.


    1) yes this is normal

    2) this is also normal

    The witnessed behaviour of visibility in Explorer has everything to do with the "caching behaviour" of Explorer itself

    You can see (but not access) the disk on a node where the disk is currently not online, but was online since the last start of Explorer (usually reboot). You can access the disk only on the node where the disk is online. You cannot see the disk on nodes which have been rebooted, but the disk has not been online since the reboot.

    HTH,

    Edwin.

     

    Thursday, May 27, 2010 9:24 AM
    Moderator

All replies

  • Hi,

     

    Please refer to this article:

     

    Event ID 1034 — Cluster Storage Functionality

    http://technet.microsoft.com/en-us/library/cc756229(WS.10).aspx

     

    Hope it helps.

     

    Tim Quan - MSFT

     

    Wednesday, February 17, 2010 6:31 AM
    Moderator
  • hi a little

    both machines report the disk as online

     but it does not seem useable

    where do match the signatue to i am using microsoft iscsi initator
    here is reoprt i get on the disk
    Listing private properties for 'Cluster Disk 2':

    Listing status for resource 'Cluster Disk 2':

     

    Resource             Group                Node            Status

    -------------------- -------------------- --------------- ------

    Cluster Disk 2       Cluster Group        V1              Online

    T  Resource             Name                           Value

    -- -------------------- ------------------------------ -----------------------

    D  Cluster Disk 2       DiskIdType                     0 (0x0)

    D  Cluster Disk 2       DiskSignature                  3203887211 (0xbef7706b)

    S  Cluster Disk 2       DiskIdGuid                    

    D  Cluster Disk 2       DiskRunChkDsk                  0 (0x0)

    B  Cluster Disk 2       DiskUniqueIds                  10 00 00 00 ... (148 bytes)

    B  Cluster Disk 2       DiskVolumeInfo                 01 00 00 00 ... (48 bytes)

    D  Cluster Disk 2       DiskArbInterval                3 (0x3)

    S  Cluster Disk 2       DiskPath                      

    D  Cluster Disk 2       DiskReload                     0 (0x0)

    D  Cluster Disk 2       MaintenanceMode                0 (0x0)

    D  Cluster Disk 2       MaxIoLatency                   1000 (0x3e8)

     

     

    Thursday, February 18, 2010 10:12 AM
  • I suspect that there is something wrong with the Disk setup in Cluster

    First of all, re-run a Full Validation, and ensure there are no errors.

    When there are no errors: Please do the following, NOT on the Quorum disk, leave the Quorum disk alone for now:
    1) Delete the disk from cluster
    2) check that in both nodes you can see the disk in Disk Management, showing as offline
    3) NODE1: Bring the Disk online in Disk Management
    4) NODE1: Ensure the disk has appropriate drive letter, and you can access the disk/files
    5) NODE1: Bring the Disk offline in Disk Management
    6) NODE2: Bring the Disk online in Disk Management
    7) NODE2: Ensure the disk has appropriate drive letter, and you can access the disk/files
    8) NODE2: Bring the Disk offline in Disk Management

    At this point you have checked basic access from the nodes to the Disk.

    9) Add the Disk to Cluster (Available Storage)
    10) Online the Disk in Cluster
    11) Note down the node which owns this disk
    12) Check that you can access this disk from this owning node
    13) on CMD prompt execute "cluster . group "Available Storage" /move"
    14) Check that you can access this disk from the other node (you just moved the ownership)

    If you can access the disk from both nodes (by moving the disk) then this is normal setup, you can now use this disk for an application/VMs

    HTH,
    Edwin.
    Thursday, February 18, 2010 12:59 PM
    Moderator
  • ok that all worked so the storage seems to be working but when i try to import a physical machine using VMM it tells me

     There are no volumes available for placement. For clustered hosts, all shared storage is currently in use or not available. In addition, some of the volumes may be marked as unavailable for placement. 

    yet i have lots a space and only vmm running on it
    which is what led me think there was a problem
    Thursday, February 18, 2010 2:35 PM
  • ok got the import working if the clustered disk and quorum is available to only one server at time is possible to set up two of each
    so two clustered disk and two quorum to use both machines more effectivitely
    Thursday, February 18, 2010 3:39 PM
  • nope the error on  import wizard is back

     There are no volumes available for placement. For clustered hosts, all shared storage is currently in use or not available. In addition, some of the volumes may be marked as unavailable for placement. 

    Thursday, February 18, 2010 4:17 PM
  • any idead why i am getting this error the drive is up and running
    VM's are on it and working
    and i can manually create a vm on it
    but the import wizards says no drives are available on both servers
    Friday, February 19, 2010 8:39 AM
  • Did this issue ever get resolved?  I'm having similar issues.  I have an IBM SAN.  I created LUN on this.  On failover cluster, I have the disk online and it is the F:\ drive.  Only one node can access this (out of 2 total nodes).  In the Failover management console, Under "Nodes" Only one has the disk show up.  I would assume both nodes should be able to access the disk at the same time.  I'm using the Microsoft iSCSI initiator to map the storage.  For the LUN I have two initiators.  Each nodes is using a separate initiator.  Any thoughts?

    Zach


    Zach Smith
    Tuesday, May 18, 2010 3:03 PM
  • I would assume both nodes should be able to access the disk at the same time. 

    No, that would be the wrong assumption.

    Normally clustered disks are only accessible by the node where they are online in the cluster. The other node(s) will not have access to the disk.

    Cluster Shared Volumes (in 2008-R2) are Mounted NTFS volumes on clustered disks, but ALL nodes have access to these volumes simultaneously.

    Hope this answers your question,

    Rgds,

    edwin.

     

    Wednesday, May 19, 2010 11:59 AM
    Moderator
  • Edwin: Sorry to bring topic related to Windoes 2003 x64 cluster.

    As per your last answer "Normally clustered disks are only accessible by the node where they are online in the cluster. The other node(s) will not have access to the disk."

    When you accessiblity, does it mean browsing, creating and reading files & folder?

    the reason I am asking you is becuase of my current scenario in my organisation

    We have a environment of 5 node cluster running on windows 2003 X64bit.

    the nodes names are
    CNDL009A
    CNDL009B
    CNDL009C
    CNDL009D
    CNDL009E

    I am able to SEE (visible) quorum disk (Disk Q:) from windows explorer and Disk management, on node CNDL009A and CNDL009D. But i am not able to see through Windows  Explorer / Disk Management on node CNDL009B, CNDL009C, CNDL009E

    My question is, 1 )whether its normal to have this type of situation.  2) Incase of Cluster group failure on CNDL009A, and when cluster group moves to CNDL009B whether IT would be able to stay there.

    my windows team says: Drive Q (quorum) will not be visible in explorer / disk management until it is or has been active on that node.

    When i say 'SEE' or 'VISIBLE' from windows explorer and DISK MGMT means I can see DRIVE Q but i cannot access and when I say Accessible, it means I can create file and browse Drive Q for the node which owns it.

    Please help me clarify. your help is very much appreciated

    Friday, May 21, 2010 5:54 PM
  • My question is, 1 )whether its normal to have this type of situation.  2) Incase of Cluster group failure on CNDL009A, and when cluster group moves to CNDL009B whether IT would be able to stay there.


    1) yes this is normal

    2) this is also normal

    The witnessed behaviour of visibility in Explorer has everything to do with the "caching behaviour" of Explorer itself

    You can see (but not access) the disk on a node where the disk is currently not online, but was online since the last start of Explorer (usually reboot). You can access the disk only on the node where the disk is online. You cannot see the disk on nodes which have been rebooted, but the disk has not been online since the reboot.

    HTH,

    Edwin.

     

    Thursday, May 27, 2010 9:24 AM
    Moderator
  • Hi Edwin,

    I wish to view quorum from both nodes simultaneously. is that possible, if yes, how?

    Regards.


    Thursday, December 5, 2013 1:10 PM
  • But when I shut down/reboot the primary NODE1 in the cluster, the share is no longer available from the Secondary NODE2. I cannot access is via admin share Clustername\ClusterStorage$ nor the NODE1\Sharename
    Friday, February 23, 2018 9:56 PM