I believe most of you tested HyperV Replica Reverse Replication. Again, there is always no issue on planned scenario. While unplanned scenario, you might notice some annoying catch. Whereby after an unplanned failover event, to re-enabled reverse replication, HyperV Replica will send a full block of VHDx across the network back to Primary Replica Server, instead of just delta changes. Surely, send a full chunks of VHDx across LAN is fine, but thinks about across the precious WAN bandwidth. It surely hurts in terms of cost and time.
So the ‘workaround’ this, HyperV Replica have a third option for initial replication, ‘use an existing virtual machine on the replica server as initial copy.’
This would force Hyper-V host to do a full blocks comparison on both ends and synchronize delta changes across network. Although it does work perfectly, do keep in mind that it is intense on Storage IOPS.