Bcachefs Linux File-System Benchmarks vs. Btrfs, EXT4, F2FS, XFS

[F.Ultra]

I'm curious about that too. According to the ArchWiki, "Unlike Btrfs and ZFS, the CRC32 checksum only applies to the metadata and not actual data." If that's no longer true, I hope someone will provide more info here.

XFS only checksums the meta-data (and you can enable this for ext4 aswell). Checksums on blocks (aka the actual data) is a world of hurt for a filesystem that is not CoW, this is one of the reasons that both ZFS and Btrfs is CoW.

[starshipeleven]

It does however give a nice handy point in time performance snapshot. and even though you are biased towards BTRFS, you have to admit that it is not a bad start!

It does barely have basic feature parity with ext4 and it still has results all over the place. We can talk of "start" when it's at least COW

[timofonic]

Michael, Y U no test ZFS?

It's available in Ubuntu.

I agree.

I don't agree with CDDL code, but it would be good to know a far more complete benchmark.

Something I dislike of Phoronix benchmarks is that they seem incomplete to me. It feels like an artificial measure to show certain results rather than others.

Michael : Why do you ask donations and subscriptions all the time but not do your homework in a proper way? I'm sure you can do it a lot better, please try it

[PuckPoltergeist]

It does barely have basic feature parity with ext4 and it still has results all over the place. We can talk of "start" when it's at least COW

From the todo-page, NOCOW still has to be done: https://bcachefs.org/Todo/

So it's neither COW nor NOCOW. How the f*** does this FS store any data?

[timofonic]

From the todo-page, NOCOW still has to be done: https://bcachefs.org/Todo/

So it's neither COW nor NOCOW. How the f*** does this FS store any data?

Fairy dust!

[jacob]

That'd be a short list. Only data checksum is implemented in XFS. And I'm not sure it's actual data checksum and not only metadata.

Thinking about it, I wonder if it's even possible to have data checksum on a non-COW filesystem.

[jacob]

XFS only checksums the meta-data (and you can enable this for ext4 aswell). Checksums on blocks (aka the actual data) is a world of hurt for a filesystem that is not CoW, this is one of the reasons that both ZFS and Btrfs is CoW.

I think it's the other way around. ZFS and BTRFS have CoW as one of their main design features, and incidentally it allowed them to support data checksum.

[PuckPoltergeist]

Thinking about it, I wonder if it's even possible to have data checksum on a non-COW filesystem.

Why should this be impossible on NOCOW filesystems?

[AndyChow]

By now, it's evident that BTRFS is badly designed. Bcachefs might not have all the features, but they are building them slowly, in a sane way, once the basics have been mastered and work. With BTRFS, everything was thrown together, and it doesn't work. A couple of weeks ago, a pcie hardware failure caused my system to require a hard reboot. My raid-1 btrfs had the last leaf on the most recent tree corrupted. There was absolutely no way to repair it. The only thing I could do was dump the files in another array, and destroy and fe-format the btrfs array. All attempts to recover and all commands were done by a btrfs developer.

So with BTRFS, a raid-1 array that has the very last block of the very last write broken due to a power failure corrupts the entire filesystem. And there is no way to recover. So how is btrfs even COW? My understanding of COW is that you could just truncate the last modifications and recover everything not too new. But no, it doesn't, not with BTRFS.

[AndyChow]

Why should this be impossible on NOCOW filesystems?

It's not impossible. You can do it manually with hashdeep. I guess it would be hard to do it in "real-time", like it's done in COW, but you could have it for any file not being in the process of being modified.