locked
Database disconnected and resynchronising after DAG member activation RRS feed

  • Question

  • Hi,

    We have 3 mailbox servers in a DAG called MX1 MX2 and MX6, each hosting a copy of all of our databases. MX1 and 2 are at our HQ, and MX6 is at a second site. I activated a couple of database copies on MX2 (MX1 was previously holding all the active copies). The copy of these databases are now Healthy on MX1, but on MX6 they are reporting "disconnected and resynchronising".

    If I suspend and activate them, they still report "disconnected and resynchronising".

    I looked through the error logs and found:

    The Microsoft Exchange Replication service was unable to perform an incremental reseed of database copy 'S2_A\MX6' due to a network error. The database copy status will be set to Disconnected. Error A timeout occurred while communicating with server 'MX2'. Error: "A connection could not be completed within 5 seconds.

    This is the same for the other database.

    Running "test-replicationhealth" gives:

    Server           : MX6
    Check            : DBDisconnected
    CheckDescription : Checks if any database copies are in the 'DisconnectedAndHealthy' state.
    Result           : *FAILED*
    Error: Failures:

    Continuous Replication for database 'R1\MX6' is in a 'DisconnectedAndResynchronizing' state on server 'MX6'. Communications couldn't be established with the server hosting the active copy of this database. Log shipping is currently stopped for this database copy.

    Continuous Replication for database 'S2_A\MX6' is in a 'DisconnectedAndResynchronizing' state on server 'MX6'. Communications couldn't be established with the server hosting the active copy of this database. Log shipping is currently stopped for this database copy.

    Wednesday, November 7, 2012 11:33 AM

Answers

  • I think I have found the problem.

    It looks like the network architects have configured static routes between MX6 and the other exchange servers. MX2 is a replacement and has a different IP address for the DAG replication to the server which it recently replaced. We have conducted a few tests and this looks like the fault. Reactivating the databases on MX1 let the databases re-activate on MX6.


    • Marked as answer by snorri788 Wednesday, November 7, 2012 1:39 PM
    • Edited by snorri788 Wednesday, November 7, 2012 1:39 PM
    Wednesday, November 7, 2012 1:39 PM

All replies

  • Hi

    How large are the databases and how fast is your connection to the second site?  The best way to resolve this is to reseed the databases to MX6 from MX2.

    Cheers, Steve

    Wednesday, November 7, 2012 11:56 AM
  • I think I have found the problem.

    It looks like the network architects have configured static routes between MX6 and the other exchange servers. MX2 is a replacement and has a different IP address for the DAG replication to the server which it recently replaced. We have conducted a few tests and this looks like the fault. Reactivating the databases on MX1 let the databases re-activate on MX6.


    • Marked as answer by snorri788 Wednesday, November 7, 2012 1:39 PM
    • Edited by snorri788 Wednesday, November 7, 2012 1:39 PM
    Wednesday, November 7, 2012 1:39 PM