ZFS 2.2.0 (RC): Block Cloning merged

Also see this issue: <a href="https:&#x2F;&#x2F;github.com&#x2F;openzfs&#x2F;zfs&#x2F;issues&#x2F;405">https:&#x2F;&#x2F;github.com&#x2F;openzfs&#x2F;zfs&#x2F;issues&#x2F;405</a><p>&gt; It is in FreeBSD main branch now, but disabled by default just to be safe till after 14.0 released, where it will be included. Can be enabled with loader tunable there.<p>&gt; more code is needed on the ZFS side for Linux integration. A few people are looking at it AFAIK.

I also see, VERY interesting:<p><pre><code> - Linux container support (So overlayfs is ZFS aware) - BLAKE3 checksums (So faster checksums) - Zstd early abort (I though Zstd already had early abort! This means you can enable Zstd compression by default, and if the data isn&#x27;t compressible, it aborts the operation and stores it uncompressed, avoiding the extra CPU use. - Fully adaptive ARC eviction (Separate adaptive ARC for data and for metadata. Until now metadata ARC was almost FIFO and not efficient) - Prefetch improvements (That&#x27;s always nice)</code></pre>

A M A Z I N G<p>Have been looking forward to this for years!<p>This is so much better than automatically doing dedup and the RAM overhead that entails.<p>Doing offline&#x2F;RAM+in memory dedup size optimizations seem like a really good optimization path. In the spirit of also paying only what you use and not the rest.<p>Edit: What&#x27;s the RAM overhead of this? Is it ~64B per 128kB deduped block or what&#x27;s the magnitude of things?

Anyone here using ZFS in production these days? If so what OS and implementation? What have been your experiences or gotchas you experienced?

What are applications that benefit from block cloning?

&gt; Note: currently it is not possible to clone blocks between encrypted datasets, even if those datasets use the same encryption key (this includes snapshots of encrypted datasets). Cloning blocks between datasets that use the same keys should be possible and should be implemented in the future.<p>Once this is ready, I am going to subdivide my user homedir much more than it already is. The biggest obstacle in the way of this has been that it would waste a bunch of space until the snapshots were done rolling over, which for me is a long time (I keep weekly snapshots of my homedir for a year).

Finally! This is exactly what I want for handling photos and media on my home server. We have phone&#x27;s dumping their camera rolls onto the machine via Syncthing, but the easiest way to share content safely is just to make copies of the file and go &quot;storage is cheap&quot;. But the reality has always been closer to &quot;storage should handle deduplication&quot; - with this feature landing that can finally be a reality.<p>EDIT: Not to mention, this should make ZFS best-in-class for handling container filesystems. Deduplicating all the files in the individual layers means you can pretty much dispense with awkwardly trying to maintain inheritance trees.

Does ZFS or any other FS offer special operations which DB engine like RocksDB, SQLite or PostgreSQL could benefit from if they decided to target that FS specifically?

Wow, I was under the impression that this had long been implemented already (as it&#x27;s already in btrfs and other commercial file systems)<p>Awesome!

Filesystem-level de-duplication is scary as hell as a concept, but also sounds amazing, especially doing it at copy-time so you don&#x27;t have to opt-in to scanning to deduplicate. Is this common in filesystems? Or is ZFS striking out new ground here? I&#x27;m not really an under-the-OS-hood kinda guy.

Technically, yes: through the use of Truenas that gives us API access to iscsi on ZFS.

Is the overlayfs&#x2F;ZFS integration first-class? (i.e. COW on overlayfs is implemented at block layer on ZFS?)

Pardon my ignorance, but isn’t this same as btrfs snapshots? Any advantage of using zfs over btrfs?

Do Btrfs or ext4 offer this?

Didn&#x27;t btrfs have this like a ~decade ago?