none
Testing DR

    Question

  • Hi All,

    Some background, we have a primary site and a DR site. Primary site has 2 mailbox servers and 2 cas servers. Our DR site has 1 mailbox server and 1 cas server. This year they want to only cut the link between the 2 sites. 

    The idea is to create a DR test that would allow our primary site to remain up and functional but allow us to test in the DR site. 
    So the flow would be 
    Cut the link between the sites
    take snapshots of VM's (DC's, Exchange, SQL..etc) 
    Do testing in DR
    Rollback snapshots
    reconnect link between sites
    everything syncs back up

    Now my issue with this is that Exchange will still be mounted and running in our primary site. I will have to force the databases online in DR. What is the best way to do this?
    I was thinking that I could just run move-ActiveMailboxDatabase -server Primbx01 -ActiveOnServer DRMbx01 -SkipActiveCopyChecks. I tested this in a lab and it did not mount the databases.

    Tuesday, May 1, 2018 4:18 PM

Answers

  • This is a little old but I was able to get it working.

    I used this site as reference. My scenario: 2 sites (Primary and DR). In the primary site we have 2 cas and 2 MBX. In the DR site we have 1 cas and 1 mbx. We have a 3 node DAG. We took snapshots of the DC's and Exchange (application consistent). Then we cut the link between the 2 sites. Exchange stayed up and operational in the primary site. We were then able to force the DR site online and do testing. Once everything was tested and signed off on, we shutdown Exchange and rolled back the snapshots on the DR nodes. We rolled back the snapshots on the DR DC and brought everything back online and established the link between the 2 sites. The DR DAG node started replicating as soon as the link was re-established and was synced back up in less than 5 minutes.

    • Marked as answer by TRDx2 Wednesday, May 16, 2018 2:10 PM
    Wednesday, May 16, 2018 2:07 PM

All replies

  • Hi TRDx2,

    As far as I know, making snapshot for Exchange is not supported, using the snapshots can have unintended and unexpected consequences for a server application that maintains state data, such as Exchange. Details see:

    Exchange 2013 virtualization


    Best Regards,
    Niko Cheng


    Please remember to mark the replies as answers if they helped. If you have feedback for TechNet Subscriber Support, contact tnsf@microsoft.com.


    Click here to learn more. Visit the dedicated forum to share, explore and talk to experts about Microsoft Teams.

    Wednesday, May 2, 2018 7:38 AM
    Moderator
  • I don't have much of a say in how we test our DR. I have read several posts where others are testing there DR in similar fashion. My question relates to how do I bring the single node online in our DR site. Can anyone provide insight on this? Here is one such post but it relates to Exchange 2010. I will be trying to test this on our lab tomorrow and would appreciate any guidance.


    Wednesday, May 2, 2018 8:14 PM
  • I don't have much of a say in how we test our DR. I have read several posts where others are testing there DR in similar fashion. My question relates to how do I bring the single node online in our DR site. Can anyone provide insight on this? Here is one such post but it relates to Exchange 2010. I will be trying to test this on our lab tomorrow and would appreciate any guidance.


    Is this a DAG? I'm confused why you want to test with a snapshot if so.

    If this isn't a DAG, then I'm confused why you want to test with snapshot as well :) 

    If not a DAG, then you should be using just the Exchange aware backups and using Database portability to test.

    https://technet.microsoft.com/en-us/library/dd876926(v=exchg.150).aspx

    Wednesday, May 2, 2018 10:01 PM
    Moderator
  • I believe it's supported if you dismount the databases and/or stop the Information Store service first.  I'd shut down all Exchange services first.

    Ed Crowley MVP "There are seldom good technological solutions to behavioral problems."
    Celebrating 20 years of providing Exchange peer support!

    Monday, May 7, 2018 10:08 PM
    Moderator
  • This is a little old but I was able to get it working.

    I used this site as reference. My scenario: 2 sites (Primary and DR). In the primary site we have 2 cas and 2 MBX. In the DR site we have 1 cas and 1 mbx. We have a 3 node DAG. We took snapshots of the DC's and Exchange (application consistent). Then we cut the link between the 2 sites. Exchange stayed up and operational in the primary site. We were then able to force the DR site online and do testing. Once everything was tested and signed off on, we shutdown Exchange and rolled back the snapshots on the DR nodes. We rolled back the snapshots on the DR DC and brought everything back online and established the link between the 2 sites. The DR DAG node started replicating as soon as the link was re-established and was synced back up in less than 5 minutes.

    • Marked as answer by TRDx2 Wednesday, May 16, 2018 2:10 PM
    Wednesday, May 16, 2018 2:07 PM
  • This is a little old but I was able to get it working.

    I used this site as reference. My scenario: 2 sites (Primary and DR). In the primary site we have 2 cas and 2 MBX. In the DR site we have 1 cas and 1 mbx. We have a 3 node DAG. We took snapshots of the DC's and Exchange (application consistent). Then we cut the link between the 2 sites. Exchange stayed up and operational in the primary site. We were then able to force the DR site online and do testing. Once everything was tested and signed off on, we shutdown Exchange and rolled back the snapshots on the DR nodes. We rolled back the snapshots on the DR DC and brought everything back online and established the link between the 2 sites. The DR DAG node started replicating as soon as the link was re-established and was synced back up in less than 5 minutes.

    Not sure what you were testing if using a DAG.
    Thursday, May 17, 2018 12:42 AM
    Moderator
  • This is for a DR site test. We are creating a DR bubble. We are isolating our DR site and bringing all our production systems online to test within that bubble while leaving our production site up and operational. Once all systems have been validated we will role back to just prior to the site to site link being cut and let all DAG and AG nodes sync back up.  I know this is not how DAG's and AG's were designed to be used but sometimes we have to work with the directives given. 
    Thursday, May 17, 2018 1:42 AM