none
DPM 2007 SP1 Exchange 2007 SP2 RU4 CCR cluster backup issue RRS feed

  • Question

  • I am backing up an exchange server with multiple storage groups. All backs up fine except for one storage group.   

    It gets a consistency error: The replica of Storage group mbx2sg12 on MAILBOX2.XMBXCD.nrel.gov is inconsistent with the protected data source. All protection activities for data source will fail until the replica is synchronized with consistency check. You can recover data from existing recovery points, but new recovery points cannot be created until the replica is consistent.

    For SharePoint farm, recovery points will continue getting created with the databases that are consistent. To backup inconsistent databases, run a consistency check on the farm. (ID 3106)

    Data consistency verification check failed for mbx2sg12db.edb of Storage group mbx2sg12 on MAILBOX2.XMBXCD.nrel.gov. (ID 30146 Details: Unknown error (0xfffffb4a) (0xFFFFFB4A))

    If I turn off integrity checking the backup succeeds. If I turn it back on it fails again. Looking for a solution so I can keep integrity checking turned on.

    Monday, August 30, 2010 9:38 PM

Answers

  • Hi,

     

    The eseutil.exe that runs on the DPM server after every backup is detecting database corruption.

    0xFFFFFB4A = JET_errDatabaseCorrupted

    You will need to make repairs to that DB on the production server, or restore a previous version of the DB before the corruption occured.


    Regards, Mike J [MSFT] This posting is provided "AS IS" with no warranties, and confers no rights.
    Monday, August 30, 2010 10:12 PM
    Moderator

All replies

  • Hi,

     

    The eseutil.exe that runs on the DPM server after every backup is detecting database corruption.

    0xFFFFFB4A = JET_errDatabaseCorrupted

    You will need to make repairs to that DB on the production server, or restore a previous version of the DB before the corruption occured.


    Regards, Mike J [MSFT] This posting is provided "AS IS" with no warranties, and confers no rights.
    Monday, August 30, 2010 10:12 PM
    Moderator
  • I have a similar issue but the failures are intermittent.  I have a CCR cluster and I have done these steps:

    On the active node:
    1. Suspend SG replication
    2. Dismount the DB
    3. Move all log files for the SG to another location
    4. Mount the DB
    5. On the passive node, restart replication with update-storagegroupcopy -identity <sg-identity> -deleteexistingfiles

    That did not help so I stopped replication again, dismounted the DB and ran: isinteg -s "<cms-name>" -fix -test alltests
    Then mount the db and repeat step 5 to restart replication

    Even after that, I had 5 SG that would not back up

    So after a couple days I rebooted the standby node of the cluster and ran "Perform Consistency Check" and all but two of the SGs will successfully recreate a restore point / Express full backup,

    If I uncheck the box to "Run Eseutil to check data integrity" and it works, would that verify a corrupt DB?

    What other steps can be taken to check the Db for corruption?

    The two that will not back up are large-ish (40GB) but another DB that is 45GB does back up fine (it was one of the five that were failing until the reboot)
    Could I not have enough space on the database drive to do the vss work needed for the backup?

    Thanks!

    Friday, September 17, 2010 4:28 AM