locked
Exchange 2010 DAG - Best Practises when Disabling Circular logging RRS feed

  • Question

  • What is the best process to following when changing the circular logging setting on Databases within a DAG?

    We have 2 MB servers in the DAG. From what I can see, after the setting is changed, a dis-mount and re-mount is required for each Database.

    Should I simply dismount and remount them one by one, or is there a script that should be run first to eliminate downtime?

    Do I need to suspend the database copy first?

    Also, how can you confirm that the change took effect after the re-mount? When I tested disabling circular logging, the 'get-mailboxdatabase | fl command' showed that it was disabled immediately after unchecking the checkbox, even before re-mounting the database?


    DQ
    Thursday, July 14, 2011 4:29 PM

Answers

  • Hi DQ,

    I don't think there is a best practice on this special scenario. In my opinion, you'd better disable circular logging for all DBs, and dismount\mount them one by one. And I think you don't have to stop the DB copy. The DB'S circular logging status is stored in AD with a name of msExchESEParamCicularLog, 1 for enable and 0 or disable. So no matter the active copy on which server, the status is always read from AD and related to DB, not the copies.

    In my test, I did what I said, nothing wrong.

    Hope it can help.


    Jack Zhou
    • Marked as answer by dq72 Wednesday, July 27, 2011 3:27 PM
    Wednesday, July 27, 2011 6:18 AM

All replies

  • Enabling/Disabling circular logging is directly related with backup with Exchange. mount/dismount is recommended to let IS service let inform for this changes immedeatly, But as you said it work without do it also.

    Exchange Circular Logging and VSS Backups:

    http://blogs.technet.com/b/exchange/archive/2010/08/18/3410672.aspx


    Anil MCC 2011,ITIL V3,MCSA 2003,MCTS 2010, My Blog : http://messagingschool.wordpress.com
    • Proposed as answer by Xiu Zhang Friday, July 15, 2011 7:32 AM
    • Marked as answer by Xiu Zhang Monday, July 18, 2011 6:06 AM
    • Unmarked as answer by dq72 Monday, July 18, 2011 3:43 PM
    • Unproposed as answer by dq72 Monday, July 25, 2011 3:44 PM
    Thursday, July 14, 2011 4:43 PM
  • Still not clear on this question: "How can you confirm that the change took effect after the re-mount? When I tested disabling circular logging, the 'get-mailboxdatabase | fl command' showed that it was disabled immediately after unchecking the checkbox, even before re-mounting the database?" If this command simply shows the setting in the gui, however a mount\dismount is required how can it be confirmed as active?

     

    Also, "Do I need to suspend the database copy first?" or just dismount and mount?
    Thanks,



    DQ
    Monday, July 18, 2011 3:47 PM
  • Hi,

     

    When you combine circular logging with continuous replication, you have a new type of circular logging called continuous replication circular logging (CRCL), which is different from ESE circular logging. Whereas ESE circular logging is performed and managed by the Microsoft Exchange Information Store service, CRCL is performed and managed by the Microsoft Exchange Replication Service.

     

    Specifically, the Microsoft Exchange Replication Service manages CRCL so that log continuity is maintained and logs are not deleted if they are still needed for replication. The Microsoft Exchange Replication Service and the Microsoft Exchange Information Store service communicate by using remote procedure calls (RPCs) regarding which log files can be deleted.

     

    Note: Circular logging must not be enabled for the specified mailbox database when you want to add database copy.

     

    So you may try to restart Microsoft Exchange replication services.

     

    Regards,

    Xiu

    • Proposed as answer by Xiu Zhang Wednesday, July 20, 2011 5:37 AM
    • Marked as answer by Xiu Zhang Friday, July 22, 2011 7:12 AM
    • Unmarked as answer by dq72 Monday, July 25, 2011 3:34 PM
    • Unproposed as answer by dq72 Monday, July 25, 2011 3:43 PM
    Tuesday, July 19, 2011 2:55 AM
  • Hi,

     

    When you combine circular logging with continuous replication, you have a new type of circular logging called continuous replication circular logging (CRCL), which is different from ESE circular logging. Whereas ESE circular logging is performed and managed by the Microsoft Exchange Information Store service, CRCL is performed and managed by the Microsoft Exchange Replication Service.

     

    Specifically, the Microsoft Exchange Replication Service manages CRCL so that log continuity is maintained and logs are not deleted if they are still needed for replication. The Microsoft Exchange Replication Service and the Microsoft Exchange Information Store service communicate by using remote procedure calls (RPCs) regarding which log files can be deleted.

     

    Note: Circular logging must not be enabled for the specified mailbox database when you want to add database copy.

     

    So you may try to restart Microsoft Exchange replication services.

     

    Regards,

    Xiu


    Thanks for the reply Xiu, but I had read the Microsoft documents with the explanation.

    What I am looking for is the recommended (best practise) steps to follow when disabling circular logging in a production environment with 4 DB's and 4 copies in a DAG. I want to minimize downtime. This is required so that I can initiate incremental backups with our Symantec backup product.

    As an example of the answer I am looking for, note the following suggestions:

                     : Disable circular logging and then reboot 1 DB server, once it comes up reboot the second?

               Or, : Stop the DB copy, disable circular logging, resume DB copy

               Or, : Run the startdagservermaintenance script, disable circular logging, run the stopdagservermaintenance script.

      


    DQ
    Monday, July 25, 2011 3:43 PM
  • Hi DQ,

    I don't think there is a best practice on this special scenario. In my opinion, you'd better disable circular logging for all DBs, and dismount\mount them one by one. And I think you don't have to stop the DB copy. The DB'S circular logging status is stored in AD with a name of msExchESEParamCicularLog, 1 for enable and 0 or disable. So no matter the active copy on which server, the status is always read from AD and related to DB, not the copies.

    In my test, I did what I said, nothing wrong.

    Hope it can help.


    Jack Zhou
    • Marked as answer by dq72 Wednesday, July 27, 2011 3:27 PM
    Wednesday, July 27, 2011 6:18 AM
  • Hi DQ,

    I don't think there is a best practice on this special scenario. In my opinion, you'd better disable circular logging for all DBs, and dismount\mount them one by one. And I think you don't have to stop the DB copy. The DB'S circular logging status is stored in AD with a name of msExchESEParamCicularLog, 1 for enable and 0 or disable. So no matter the active copy on which server, the status is always read from AD and related to DB, not the copies.

    In my test, I did what I said, nothing wrong.

    Hope it can help.


    Jack Zhou


    Thanks Jack,  even though not directly listed step by step in any documentation that I could find, this is what I came up with as well. A second set of eyes always helps, and I'll give it a go.

    Thanks again,

    Darrell


    DQ
    Wednesday, July 27, 2011 3:29 PM
  • Hi Darrell,

    I believe that this, from Scott Schnoll, might be a more definitive description of the process:

     

    • If the mailbox database is not replicated, it will use JET circular logging. In this case, enabling or disabling JET circular logging will require a dismount and mount of the database.

     

    • If the mailbox database is replicated, it will use CRCL. In this case, enabling or disabling CRCL takes effect dynamically; there is no need to dismount and re-mount the database.

     

    From http://blogs.technet.com/b/scottschnoll/archive/2011/06/27/circular-logging-and-mailbox-database-copies.aspx

    Hope this is of use.

    Best Regards

    Monday, November 14, 2011 3:19 PM