WAN Accelerator & ReFS

[edmorcos]

Hello, I have a few architecture questions about ReFS format and WAN Accelerator. Please delete this post if it is considered as "support questions".

ReFS:
We have 2 Standalone hosts running Windows Server 2016 with 8 hyper-v backed-up from Server A to Server B every night (about 2.5Tb)
The same vms are also replicated from Server A to Server B every 20mins.

Server A has one volume in NTFS format that hold original hyper-vs (production).
Server B has a newly created volume in ReFS 64kb format that serves as a repo.

Backup jobs are not using Fast Cloning. Does both servers need to have ReFS volumes for it to work? Is it worth it for such a small size data do switch the production Server A from NTFS to ReFS?

WAN Accelerator:

Connection with remote site is only 4Mbps. Before, we use to replicate two of the main hypervs (W.S. replication) every 5mins and it was working fine.
The goal is to to keep an offsite copy of our data.

Now, we want to considering to take advantage of WAN Accelerator and have copies of all VMs off-site to Server C.

- Do you recommend using a Backup Copy job or Replication keeping in mind the connection speed?
- Does it help to have Server C with ReFS Volume or it doesn't matter?
- Any special recommendations / tips?

Thanks

[foggy]

1. The source server doesn't need to be ReFS formatted for FastCLone to be used on target. I'd recommend reviewing the list of requirements and limitations and in case nothing applies, ask our support to review your setup closer.
2. Backup copy and replication jobs will transfer similar amount of changes, so your decision should be based on other factors, depending on what do you want to achieve with this offsite copy of your data. Whether or not to format that server as ReFS also doesn't matter in the context of WAN acceleration.