0

We have 2 VM's running Windows 2012 (not R2) in a Windows 2012 R2 Hyper-V Environment. The servers are both on the same host and are both replicated to the same server at a remote site. Both servers have the same service installed, a business application that retrieves a message from a remote MSMQ, loads the message into the local MSMQ, runs an in memory process (only disk change might be a few kb of log files), and then moves the message back to the remote server. Both servers are processing roughly the same amount of payloads.

Both servers are set to replicate every 15 minutes.

When replicating, the first server is tracking changes and replicating around 100MB every cycle, while the other server tracks and replicates 1-5GB (yep, GB).

If we stop the business process on the 2nd server, the changes drop to a few MB, so there is definitely something about how the business process is running on the 2nd server.

Both servers have the same size drives, even if we send all of the load to the 1st server, it still only generates 150MB to replicate each cycle.

Is there anything in Hyper-V that will help track down what has changed on the disk, or what is changing the disk so we can work out why there is such a discrepancy in replication size?

3
  • I do not understand, how the servers are replicated? What is the data mover?
    – A.Newgate
    Apr 17, 2018 at 11:54
  • @A.Newgate it's using Failover Clustering to replicate to DR
    – Greg
    Apr 17, 2018 at 20:43
  • All - We have also tried following this TechNet article: gallery.technet.microsoft.com/Hyper-V-Replica-Identify-f09763b6 This has proven to show a local output of 102mb within 30 minutes. Needless to say, this is confusing to us because the Hyper-V HRL file was 2.83GB at the time.
    – Dave
    Apr 18, 2018 at 23:53

0

You must log in to answer this question.