locked
SCR site shutdown RRS feed

  • Question

  • Hi

    We have an Exchange 2007 CCR cluster in our office, along with Hub/CAS servers. We also have a DR site that holds the SCR target for the CMS and Hub/CAS servers.

    The DR site will be shutdown for a day due to urgent maintenance works.

    Bit of a strange situatoin as we'd always planned this around the Prod site being unexpectedly shut down and us having to fail over the cluster to the SCR target! Given that it's the SCR target that will be unavailable, what are the best options?

    1. Suspend-StorageGroupCopy before the work is due to start and then Resume-StorageGroupCopy

    or

    2. Suspend-StorageGroupCopy before the work is due to start and then Reseed?

    Given that the failover site will be down for an entire day, I'm not sure suspend/resume would work? That's a hell of a lot of logs to have to replay?

    Tuesday, June 7, 2011 9:42 PM

Answers

  • Technically, neither are necessary.  If the target is down or inaccessible, replication will simply back up.  Regarding #2, a reseed isn't normally necessary after a suspend, so that option doesn't make any sense.  If you have enough space in your transaction log volumes to support #1, then I recommend that you do that.  Suspending for a planned outage is the best operational practice.

    One day's volume of log files should be substantially less than the size of the databases, no?


    Ed Crowley MVP "There are seldom good technological solutions to behavioral problems."
    • Proposed as answer by Ed CrowleyMVP Sunday, June 12, 2011 11:13 PM
    • Marked as answer by Novak Wu Thursday, June 23, 2011 1:18 AM
    Tuesday, June 7, 2011 10:27 PM

All replies

  • Technically, neither are necessary.  If the target is down or inaccessible, replication will simply back up.  Regarding #2, a reseed isn't normally necessary after a suspend, so that option doesn't make any sense.  If you have enough space in your transaction log volumes to support #1, then I recommend that you do that.  Suspending for a planned outage is the best operational practice.

    One day's volume of log files should be substantially less than the size of the databases, no?


    Ed Crowley MVP "There are seldom good technological solutions to behavioral problems."
    • Proposed as answer by Ed CrowleyMVP Sunday, June 12, 2011 11:13 PM
    • Marked as answer by Novak Wu Thursday, June 23, 2011 1:18 AM
    Tuesday, June 7, 2011 10:27 PM
  • Hi Ed

    Thanks for replying, makes sense.

    " If you have enough space in your transaction log volumes to support #1, then I recommend that you do that."

    I assume the above comment refers to transaction log volumes on the SCR source, as this is where the transaction logs will build up?

    Wednesday, June 8, 2011 7:04 AM
  • Yes. If there are a lot of available volumes on SCR source, you can just resume to replay the logs to the target. It’s not recommended to do a full reseed.

     

    Thanks.

    Novak


    Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.
    Friday, June 10, 2011 2:05 AM