BTRFS (BeTaRaid File-system) has been around longer than openZFS but still can't be considered a serious option for a NAS because horrible corruption bugs that plague raid 5 & 6it is more likely at this point that openZFS somehow is included in the kernel before btrfs becomes usable?
Am I the only one that calls it butterface when I read it out loud?
Usecase for raid 5 & 6?
Not sure why /g/ now wants to talk about it other than some weird contrarian point of view.RedHat, the driving force behind it, has basically told everyone to stop using it in prod and has taking it back to the development stage due to data loss issues.>>106979217you want to keep data up and available if a drive fails? duh
I wish it was in kernel, even though it's just like 2 lines to get it (was shocked actually)R5/R6 in the ZFSin Windows fork is probably more stable too, and bcache is 5+ years away
>>106978768Dunno. I don't need RAID, so been happily using btrfs for years.When I last tried ZFS on Gentoo (late 2010s), it was kind of an awkward experience cause Linux tools didn't tend to support it, even showing basic info required bespoke tools. Having lz4 compression was cool tho.
>>106978768The anti btrfs propaganda gotta be a psyop.That is if we ignore brainlets with broken discs who think checksumming is what "broke" them.That being said, ZFS is jewish oracle malware due to its license and Linus deserves to get his legs broken for how the bcachefs situation played out.
>>106979795>ZFS is jewish oracle malware due to its licensea markdown file with legal information is malware?seek help
>>106980150> "what damage can a license possibly cause? it's just a bunch of letters"
>>106979795It's just a normal copyleft license.
>>106979240>due to data loss issuesThere's also the fact they're heavily invested into XFS and made their own volume management layer for it.>>106979638I can't mount ZFS partitions through the file manager and I can't see the datasets in the disk utility (though I'm pretty sure that was the case with Btrfs too). Can't say this impacted me much. I'd say the lack of bootloader/installer support is a more major pain point.>>106978768>it is more likely at this point that openZFS somehow is included in the kernelThe only remotely likely scenario that I can envision is once Oracle stops milking Solaris and puts it to rest, they might decide to focus on giving first class support to ZFS on Oracle Linux / UEK.
>>106981403>I'd say the lack of bootloadercheck out zfsbootmenu.
>>106979795why didn't the other dogs kick his ass? wtf.
btrfs is the systemd of filesystems, dont believe me go look at their docs, its practicly an operating system in itself
>>106981403>There's also the fact they're heavily invested into XFSpeople ignore this fact for some reason. given how many XFS devs now work for IBM, it should be kind of obvious there were some incentives there. I don't blame them though. You can't just support every possible kernel config for 10 years without cutting the fat somewhere. Shame though. IMO I think Btrfs better fits the enterprise use cases better than XFS, by a lot. device mapper shit is giga bloat.
>>106978768openZFS had some kind of licensing issue so I don't think it'll ever be added to the kernel. Oracle can (and probably will) sue if it's added.
>>106979795bcachefs sucks dick, not sure why there's controversy about it
>>106983547above average programmers who happen to capture that sort of manic energy usually get a cult following of retards who promote their works, deserved or otherwise. Kind of like how, "Um Google does this, so we have to" and why I still suffer with shit like gtest and protobuf to this day.thankfully he pissed off enough idiots above him so his meme fs might finally die.
>>106979240>>you want to keep data up and available if a drive fails? duhRaid 5 becomes unreadable if one drive fails.
>>106983597That's not even a good troll so you must be retarded.
>>106983597>>106983615maybe on btrfs's implementation of raid since its systemd for disks
>>106979217you're too cheap to run proper RAID 10 or n-copy array.
>>106978768We'll get a good FS whenever the chuds at Oracle (never) decide to remove their shitty license from ZFS.Meanwhile, ext4 just works.
>>106983461It's what I use already. I'm just saying it would be nice to have more distros offer root on ZFS out of the box. I'm not sure if Ubuntu still does it (last I heard it was only offered by the server flavour installer).
>>106983808>Meanwhile, ext4 just works.Shouldn't everybody be using XFS instead?
>>106979795>nigger dog vs white dognow remove dog
>>106979795kent overstreet is a fucking moron. bcachefs is written by a retard. i don't want to trust my data to a retard.
>>106979795btrfs had bugs that would destroy the file system if you do as much as try to mount it on the wrong kernel version.
>>106983808use case for zfs when btrfs is already there?
>>106983532They haven't sued Canonical and Canonical has been packaging OpenZFS as an actual package for a while now
>>106986087you can distribute the package, you just cannot include zfs in the linux kernel and distribute the result
>>106984098Depends.Need reflinks -> XFS.Need to minimize write amplification -> XFS.Need to minimize read amplification -> ext4.
>>106979795it's a psyop to make goyim believe only the cloud and enterprise hardware can hold your data
