Bcachefs Gets "[email protected]$$" Snapshots, Still Aiming For Mainline Linux Kernel Integration

[phoronix]

Phoronix: Bcachefs Gets "[email protected]$$" Snapshots, Still Aiming For Mainline Linux Kernel Integration

Kent Overstreet who has been working relentlessly on Bcachefs for over a half-decade now issued his latest status update on this Linux file-system born out of the kernel's block cache code...

https://www.phoronix.com/scan.php?pa...-November-2021

[flower]

I still have really big hopes for bcachefs.
I will probably switch my four 12tb mirrors as soon as it's mainlained.
Erasure coding sound promising. Hope he gets it stable.

I do have good backups and i am prepared to get burned though

[pal666]

I do have good backups and i am prepared to get burned though

it sounds like you are excited about being burned

[some_canuck]

zfs is still better

[flower]

it sounds like you are excited about being burned

Well kind of. It's the price you pay for early adopting new tech. And i love tech

[intelfx]

zfs is still better

zfs cult coming out of the woodwork

[ALRBP]

I still have really big hopes for bcachefs.
I will probably switch my four 12tb mirrors as soon as it's mainlained.
Erasure coding sound promising. Hope he gets it stable.

I do have good backups and i am prepared to get burned though

The article says it would be marked experimental due to potential on disk format changes. You should definitely not use it in production until this phase is over.

[acobar]

Awesome news!

I too, as soon as it get mainlined, will put it on a spare machine and will keep making binary packages and compare the results with what is produced on my trusted ext4 partitions.

[DrYak]

Kent Overstreet (and the bcachefs fans): complain that btrsf took a decade to develop and still isn't mature in all its features.

Also Kent Overstreet: takes nearly a decade to develop bcachefs and still adding critical features now.

:-D
I am joking. I know developing filesystem is hard. And especially new-gen filesystems with modern features tend to be harderer.

Now on a more serious note:
it's good that it's progressing! I am looking forward to see how good it will eventually turn out. I also hope that his work will find some corporate sponsors and large real-world testing: probably Linux-based NAS appliances maker is going to be good candidates as bcachefs is going to be providing an interesting alternative to ZFS (simpler licensing and deployment under Linux (ZFS is legally easier to ship on BSD derivatices)) and btrfs (multitiered block caching supported as native feature (btrfs needs to rely on lvm-cache or bcache), supposedly better erasure coding once it matures (native RAID5/6 still not considered stable in btrfs - e.g. I'm using a mdadm raid underneath instead), etc)

[pgeorgi]

Kent Overstreet (and the bcachefs fans): complain that btrsf took a decade to develop and still isn't mature in all its features.

Also Kent Overstreet: takes nearly a decade to develop bcachefs and still adding critical features now.

I suppose the difference is that bcachefs is developed out of tree for a decade, while btrfs was fast-tracked into Linux upstream long before it was ready.