With btrfs’s send and receive commands you can replicate a snapshot on a different filesystem. The command operates in two modes: full and incremental. In the incremental mode, previously sent snapshots that are available on both the sending and receiving side can be used to reduce the amount of information that has to be sent to reconstruct the sent snapshot on a different filesystem; much like rsync in non-btrfs Timeshift.
It’s just a btrfs feature that Timeshift doesn’t make use of. But you could still send/receive snapshots created by Timeshift manually if you wanted to.
Btrfs feels like it will never be finished and attain “production” quality.
For me it’s still the first choice for the root filesystem though, no other in-tree filesystem has all those features.
It can be terribly slow in some instances, but with fast NVMe storage it doesn’t matter much to me tbh.
Unfortunately, it still lacks transparent encryption support, which both ext4 and f2fs already have…
It has another advantage: it works well together with systemd’s container mechanism (machinectl, systemd-nspawn).
not gain with more than 2 disks , why offers in theses conditions raid , if no benefit results in bench ?
may be raid1 if only it’s more reliable that ext4
missing repairs reliable commands in case it start wrong … , so your only use is a snapshot/timeshift and you expected that you can always restore a good snapshot …
But RAID-like is very easy with btrfs, as you don’t need an additional layer like mdraid.
You can basically just add a drive to your computer and run btrfs device add followed by a balance. It’s a bit similar to ZFS in that regard.
That’s true. It can be interpreted in a negative way (“so many bugs”) or in a positive way (“so many fixes”).
I like btrfs by default on Arch-based distros. While it is an imperfect filesystem from a stability perspective it makes rollbacks near instantaneous in the event of a problematic update. Also snapshots on btrfs are better than rsync snapshots in almost every conceivable way. Personally, I use zfs instead but that isn’t for everyone.
On the other hand, I see quite a bit less value on Fedora. System breaking updates aren’t common on Fedora and dnf already provides support for rolling back a transaction.
