Bcachefs Reining In Bugs: Test Dashboard Failures Drop By 40% Over Last Month

[phoronix]

Phoronix: Bcachefs Reigning In Bugs: Test Dashboard Failures Drop By 40% Over Last Month

As part of the latest Bcachefs fixes pull request, lead developer Kent Overstreet has provided an update on the bug situation for this advanced copy-on-write open-source file-system...

https://www.phoronix.com/news/Bcachefs-Failures-Drop-40p

[V1tol]

Well, that looks reliable.

[Mathias]

Just a little news from Drama-World:

On Kents RC5 PR, Sasha Levin tested a tool to check how many days the patches were in linux-next (results: 5x 0 days, 21 x 2 days). That was something Linus did manually earlier and complained about. While the messages had potential to result in drama, everyone kinda tried to not directly point fingers. In the end, Kent was relieved to hear >>This is not a "shame" list but rather "look a little closer" list.<< from Sasha. He concluded

I've got stuff that I hold
back for weeks (or months), and others that I'm fine with sending the
next day, once it's passed my CI.

I'm going to try to be better about talking about which patches have
risks (and why that risk is justified, else I wouldn't be sending it),
or which patches look more involved but I've got reason to be confident
about - that can get quite subtle. That's good for everyone down the
line in terms of knowing what to expect.

Linus never commented anything on this thread. Also with his complaints about Kents RC4 PR, Kent sent in the RC5 PR 6 days ahead and RC6 two days ahead of time. This is something Linus suggested as a compromise to (not) having patches in -next.

[varikonniemi]

Well, that looks reliable.

at least it has not eaten any data yet. Fixing corner case wedging that needs extending fsck is usually expected to continue for years after being declared stable, so seeing those reduce dramatically at this point is indeed very reliability inducing.

Everyone laughed at the "provocative" catchphrase but as of now it seems very possible to be the first filesystem that has never and will never eat data while being mainline.

[uid313]

Maybe he should be using Rust now since that is in the kernel?

[niner]

at least it has not eaten any data yet. Fixing corner case wedging that needs extending fsck is usually expected to continue for years after being declared stable, so seeing those reduce dramatically at this point is indeed very reliability inducing.

Can we please bury this myth?

https://www.reddit.com/r/bcachefs/comments/1akbc4e/common_traps_to_avoid_to_keep_bcachefs_from/

https://www.reddit.com/r/bcachefs/comments/1ajmj8w/automatically_recovery_from_unclean_shutdown/

https://github.com/koverstreet/bcachefs/issues/456

I mean yeah, if you define "not eaten any data" as "files don't tend to magically disappear from directories -- they show up, but then you get failures trying to stat() them" then maybe yes:

https://www.reddit.com/r/bcachefs/comments/1bluxqs/does_bcachefs_ever_losecorrupt_data_without/

[Radtraveller]

Reining, not "Reigning"

[varikonniemi]

Can we please bury this myth?

https://www.reddit.com/r/bcachefs/comments/1akbc4e/common_traps_to_avoid_to_keep_bcachefs_from/

https://www.reddit.com/r/bcachefs/comments/1ajmj8w/automatically_recovery_from_unclean_shutdown/

https://github.com/koverstreet/bcachefs/issues/456

I mean yeah, if you define "not eaten any data" as "files don't tend to magically disappear from directories -- they show up, but then you get failures trying to stat() them" then maybe yes:
https://www.reddit.com/r/bcachefs/co..._data_without/

nothing in there is a demonstration of "eating your data", mainly just speculation from users that don't know what they are doing.

As i said in my initial message, it's completely different between "a bug was hit, data is unavailable until fsck is extended to recover it" and "the data is gone / eaten" The second one has not yet happened on bcachefs since it has been mainline.

[Siuoq]

Everyone laughed at the "provocative" catchphrase but as of now it seems very possible to be the first filesystem that has never and will never eat data while being mainline.

Ext2-3-4 work great for me

[KernelCrasher]

Ext2-3-4 work great for me

I like all 3 of them except ext2 and ext3.