Hi Jeff,
Deduplication will be available on ReFS in the Windows Server, Version 1709 release, and it will function similarly to Deduplication on NTFS: https://docs.microsoft.com/en-us/windows-server/storage/data-deduplication/overview
Additionally, Deduplication will introduce a feature known as near-inline optimization, which works with ReFS mirror-accelerated parity. Writes will go directly to the mirror tier for optimal write performance. Once that data is rotated to the parity tier, this data will be deduplicated, and then ReFS will compute parity on this data. This enables ReFS + Dedup to deliver high-performance writes then dramatically reduce the data footprint when storing colder data in parity, as that data is now dedup'ed and stored in a parity resiliency scheme.
Thanks, Garrett
