exchange 2007 CCR consistency check size after reseed with DPM 2010 RRS feed

  • Question

  • Hello all, had a question regarding consistency checks. I had to reseed a couple of exchange 2007 databases after some network issues and am running a consistency check. One of the things that I'm having problems with is accurately having enough free space on either the replica or recovery point when running this consistency check. Because this is a reseed, I'll like have to check all 200gbs of a database. My question is will I need another 200gb or so of free space on either the replica or recovery point volume? I would prefer not to run out of space while trying to get these consistency checks finished.

    I understand I could just remove the protection group and create a new group much faster, but then I lose my old recovery points to my knowledge. Any info wil be appreciated!


    Wednesday, November 17, 2010 2:34 PM


All replies

  • Hi John,

    you need roughly twice the space, certainly if you want to maintain old recovery points.
    This is why the DPM-E2007 storage calculator accounts for planned CCR failovers (which is effectively same as reseeding).
    The space consumption after CCR failover or reseed occurs on the recoverypoint volume (storing the delta's), not the replica.

    Given you talk 'reseed' and out of space I take it you reseeded using Exchange are are protecting the usually passive CCR copy!?
    'Reseed' of a protected CCR copy is best done through DPM recovery of last recovery point, effectively same result but the required consistency check synchronizes way less data (Exchange reseed typically generates >90% delta churn).

    To maintain everything you need to account for indeed ~100% churn which will consume ~200GB on the recovery point volume.
    Note: this is temporary if you can afford the diskspace. After retention of recovery points that hold onto this huge churn expires these get pruned and release involve disk space (stays allocated but you problably can shrink afterwards).
    Recreating protection makes no sense because you then also create a new replica on DPM. If you cannot afford the temporary disk space to overcome it makes more sense to just delete all recovery points that will also free-up recovery point volume space and create a new recovery point now. You still lost all old recovery points but do not have to recreate the replica's (200GB each).

    You could get lucky when just deleting the latest ExpressFull (and all newer recovery points) and may free-up the big churn but keep all older recovery points. I cannot guarantee upfront this will free-up space (involves complex factors) but then again if it does not help it does not hurt either compared to alternative.
    You can use this to guide you in deleting on or more or all recovery points. This operates per data source and before or after or in between dates. Don't worry it is QA style (no params) and will clearly warn last chance abort before it changes anything at all.

    \R2 This posting is provided "AS IS" with no warranties, and confers no rights
    • Proposed as answer by Nageshbk Thursday, February 10, 2011 10:10 AM
    Sunday, November 21, 2010 12:04 PM
  • PS. if you want to learn more of what is involved in creating/deleting recovery points see this video: 
    \R2 This posting is provided "AS IS" with no warranties, and confers no rights
    Sunday, November 21, 2010 12:08 PM
  • housekeeping - closing old post.  Open a new post if you still have a need.

    Regards, A.Nadar, This posting is provided "AS IS" with no warranties, and confers no rights.

    Wednesday, June 20, 2012 9:58 PM