none
Backup failed because the database is already in the process of being backed up RRS feed

  • Question

  • Hello,

    We are in the process of moving from exchange 2003 to 2010 and I have a problem with backups on 2010.

    I’m receiving error event id 9840 once per day around 22:00 with description “backup failed because the database is already in the process of being backed up” this must be related to DPM2010 which is scheduled to run Express Full Backups at this time. I don’t seem to get any problems from incremental backups ONLY on the Full backup. Which seems to automatically correct itself 1 hour later – maybe DPM has some retry logic?

    Can someone please confirm I’ve configured the Exchange 2010 DAG / DPM environment optimally?

    We have x4 exchange servers all in single AD site; x2 configured as CAS/HT WNLB – working perfectly no problem. THEN x2 exchange server configured as mailbox role, joined together in a DAG:

    Mailbox server1 has DB1 & DB3

    Mailbox server2 has DB2 & DB4

    Within the DAG, DB1&3 are copied to server 2 AND DB2&4 copied to server 1

    DPM2010 setup = single Protection Group covering complete DAG, incremental copy every 2 hours, express full back-up once per day at 22:00.

    Initially I set the DPM incremental / full backups as per active environment; that is Server1; DB1&3 Server2; DB2&4 THEN the daily copy backups on the copy databases, so server1; DB2&4 Server2; DB1&3

    As I started to get error 9840 straight away I thought maybe I have setup DPM incorrectly and so changed incremental / full backups assigned to ALL DB’s on server1 and daily copy backups on ALL DB’s on server2. I understand DPM doesn’t care if it is working with either a live or copy database… But still I get error 9840 – not on all DB’s; on server1 the error is related to either db1 or db3, and on server2 the error is linked to either db2 or db4.

    Have I made a mistake with the way Exchange DAG is configured to load production dbs across two servers? OR a mistake with DPM and Protection Group configuration? Any idea how i can stop the errors?

    Regards

    Mark

    Thursday, September 27, 2012 12:40 PM

Answers

  • OK, I found the solution to the problem, I needed to configure the backup using two seperate Protection Groups - one per DAG node. That way you can specify Full backups start at one time on Node 1 and copy backups start at a different time on Node 2.

    Regards

    Mark

    • Marked as answer by Mark Annis Monday, October 1, 2012 11:07 AM
    Monday, October 1, 2012 11:06 AM

All replies

  • Hi

    I found a KB talking about Error 9840, perhaps you can have a look on that and try the resolution there.

    Event ID 9840 or 9607 is logged when a VSS backup operation fails in Exchange 2007 or in Exchange 2003

    http://support.microsoft.com/kb/930800

    Hope that helps

    Cheers


    Zi Feng

    TechNet Community Support

    Friday, September 28, 2012 9:08 AM
  • Thanks, I already tried that KB yesterday, BUT i still have the same issue at 22:00 again last night...

    To me it looks like DPM kicks off a full backup job on Server1 DB3 @ 22:00 (active db) and then at the same time starts a Copy backup job on Server2 DB3 (replica db). THEN one of the servers throws up an error about can't continue because another backup job is running...

    Thats why i'm wondering if i have configured my environment correctly?? What is best practice for protecting a DAG - backup of both active and replica dbs. Can the backup jobs be run at the same time on separate nodes?

    Regards

    Mark

    Friday, September 28, 2012 9:33 AM
  • Yes, you can run backups against multiple nodes against different databases. Of course, you do not backup the same database against 2 different servers. Backing up the passive node, once successful, will truncate the logs of all the database copies.

    You may want to ask this in the DPM forum to ensure you have that set up correctly:

    http://social.technet.microsoft.com/Forums/en-us/dataprotectionmanager/threads

    Friday, September 28, 2012 11:23 AM
    Moderator
  • OK, I found the solution to the problem, I needed to configure the backup using two seperate Protection Groups - one per DAG node. That way you can specify Full backups start at one time on Node 1 and copy backups start at a different time on Node 2.

    Regards

    Mark

    • Marked as answer by Mark Annis Monday, October 1, 2012 11:07 AM
    Monday, October 1, 2012 11:06 AM