I’m getting this question so much still. I’m going to answer it here (as I did in the book) to make it final.
Hyper-V Replica (HVR) replicates virtual machines from one host/cluster to another/host cluster. HVR is not physical storage replication. It doesn’t care if you use iSCSI, SMB 3.0, Fibre Channel, Storage Pools, or SAS. It doesn’t care if you use a JBOD, a SAN, internal disks or USB. It doesn’t care if you’re using CSV or simple internal NTFS. HVR is storage agnostic … it simply does not care what storage you use in the primary or the secondary sites. The storage in the primary and secondary sites can be completely different. You just need to be using storage that is supported by Windows Server 2012 Hyper-V (check the HCL).
That’s pretty definitive and there should be no remaining questions on this.
Aidan a quick question: Does deduplication work/is supported for Replica VHDs?
Haven’t looked, but I doubt it. The files are constantly active (incoming replication) so they’ll never reach the point where they are 1 day old, 6 days old, etc, for scheduled optimisation.
I have a (possibly) stupid question about Hyper-V replica: when setting up the broker one selects a default location for replicated files.
However, I cannot see where to change the replication destination for a specific VM – how does one achieve this?
It’s not per VM. It’s either per local host (1 size fits all) or per-remote replicating host/cluster.