none
Validate this cluster... - (Win 2008 r2 cluster) returning storage error (prior SQL 2008 cluster install).

    Question

  • Hello a 2 node active/passive cluster has already been setup for me & about to install SQL Server 2008 onto the active node.  But when I click on the 'Validate this cluster...' in FailOver Cluster Manager the following Storage error displays...

    List Potential Cluster Disks


      List disks visible to two or more nodes that will be validated for
      cluster compatibility. Online clustered disks will be excluded.
      Disk with id a3fb6a27 is only visible from one node. This disk
      is not eligible for cluster validation.
      Disk with id a3fb6a2f is only visible from one node. This disk
      is not eligible for cluster validation.
      Disk with id a4758624 is only visible from one node. This disk
      is not eligible for cluster validation.

      Disk with id a4758623 is only visible from one node. This disk
      is not eligible for cluster validation.

      The wrong diskette is in the drive.
      Insert %2 (Volume Serial
      Number: %3) into drive %1

      ...do I still need to run this 'Validate this cluster...' & need to address this Storage error or can this be ignored and is ready for me to install SQL 2008 cluster?  Thanks in advance.

    Thursday, July 18, 2013 9:45 PM

All replies

    • ...do I still need to run this 'Validate this cluster...' & need to address this Storage error or can this be ignored and is ready for me to install SQL 2008 cluster?  Thanks in advance.

    Hello,

    Unless and until cluster validation wizard gives you no error please dont move forward.From message above it seems SAN storage is not cluster aware and is available to only one drive.

    Simple check shutdown one node storage shoud appear completely to another node and vice versa.

    Also during validation ,run validation on both nodes .Also while runnig validation select both nodes on each machine to run validation.

    Contact ur storage team and make sure your SAN is visible to both nodes. 


    Please mark this reply as the answer or vote as helpful, as appropriate, to make it useful for other readers

    Friday, July 19, 2013 5:33 AM
  • Thanks Shanky for your reply, really appreciate it.

    When shutting down active 1st node it does look like it fails over correctly to the passive 2nd node which can also see all of the shared SAN disks in doing simple test.  I've contacted the SAN team which we're kind of scratching our heads at this moment?

    Tuesday, July 23, 2013 7:02 PM
  • Hello,

    Every disk resource needs to be presented to all nodes, they must be clustered and online.

    Use the “List all disks” on the storage test to see if all disks are visible to all nodes.
     
    Hope this helps.

    Regards,
    Alberto Morillo
    SQLCoffee.com

    Tuesday, July 23, 2013 8:41 PM
    Moderator
  • Thanks Alberto.

    The "List all disks" completely succeeds listing both nodes & all of their disks.  It looks like it's failing at the next step of "List Potential Cluster Disks" ?

    Tuesday, July 23, 2013 8:59 PM
  • Hello,

    Can you try to remove the cluster disk and re-initializing the disk via Disk Administrator?

    Regards,
    Fanny Liu


    Fanny Liu
    TechNet Community Support

    Monday, July 29, 2013 12:18 PM
    Moderator
  • Am getting this error again after what seems like I've successfully added a new shared SAN storage disk drive and then re-running cluster validation again which think I know what's happening now in that...

    A) Example Node_A storage disk's Quorum is being listed as Disk # 9 and Node_B storage disk's Quorum is listed as Disk #12...but both sizes and types are reflecting correct info.  When failed over both are reflecting same mapped 'Q' drives which is it ok if the Disk #s are off like I mention here or does every storage literal Disk # on both nodes need to match exactly and if yes then how to do that (currently have .mdfs & .ldfs on the shared drives) ?

    B) Does cluster validation even need to be re-run after what looks like to be successfully added new storage disk before proceeding to adding this newly added disk as a dependency to existing SQL Server cluster resource group?

    Thanks in advance.

    Thursday, November 21, 2013 2:02 AM