Linux 5.5 SSD RAID 0/1/5/6/10 Benchmarks Of Btrfs / EXT4 / F2FS / XFS

[oiaohm]

Oh, I didn't realise the feature got finally declared stable now. I was still remembering it as an experimental feature only.
Is it enabled by default, or is it still at the "must be configured" phase ?

https://blogs.oracle.com/linux/xfs-d...haring-reflink the link January 6, 2020 reflink is declared production ready on xfs and on by default when you create a new xfs file system using xfsprogs 5.1 and newer yes that 19 Jul 2019 https://lwn.net/Articles/794288/. Its now that getting out to distributions.

Reflinks and Filesystem meta data check-summing on by default on XFS in a growing number of distributions now, So at this point XFS is Part CoW due to Reflinks. Different way of doing copy on write. There is still the weakness of not check-summing data at this stage and you don't have your snapshots feature yet.

XFS feature list road map has quite a few still coming.

Also F2FS is log-structured, which shares some of the benefit (no-inplace overwrite, possibility to always recover by reverting to an older version, friendlier on append-mostly / overwrite averse media such as Flash, Shingeld magnetic, etc.) that CoW also provides. It is NOT checksuming its data though.

I forgot about log-structured type of file systems they are a different beast.

[S.Pam]

XFS does NOT do data checksum, so do not compare it to Btrfs or ZFS. Without data checksums a filsystem cannot guarantee the integrity of stored data.

https://en.m.wikipedia.org/wiki/Comparison_of_file_systems

[lsatenstein]

I am an xfs fan!! I do not have to do shrinking. If your xfs partition is close to maximal, beware. You can run into issues with 10% free space or less.

By the way, use of the word lastly bothers me. Something can be last, but then should lastly should mean, not last or finally, but close to last. If close to the last, how close?
Just as last is a bother, in and into bother me. Put something into the system. Where is the something? It is in the system.

Back to xfs and tests. Would data centers be using terrabyte SSDs or Spinners? Raid on disk may favour btrfs.

[oiaohm]

Back to xfs and tests. Would data centers be using terrabyte SSDs or Spinners? Raid on disk may favour btrfs.

https://docs.microsoft.com/en-us/windows-server/storage/storage-spaces/choosing-drives

This is not a Linux site but the 7 layouts Microsoft writes up there you see in data centre. Data centre can be using like 18TB spinnig(HDD). Your back ups and cold storage don't need SSD speed.

https://xfs.org/index.php/Shrinking_Support

The issue with shrinking XFS is really developers having time to implement the missing features on the Linux version of xfs. Xfs got really badly limited when it got ported to LInux.

[Yoshi]

Michael
Thanks for the test. Is it possible, in the future, to write an Overview about Gluster and Ceph? I find it really hard to get into this stuff. Maybe it is something worth looking at.

[profoundWHALE]

Obvious troll is obvious. You can do better than that...

Not a troll. I've tried btrfs 3 times and had terrible experiences each time. The first time was a long time ago, like 2010 and my system lasted a few weeks before I had to reinstall. The second time was in 2013 and I had issues with file corruption, then in 2016 I tried a RAID10 set-up with 24TB worth of drives and the only reason why I didn't lose everything was because I had everything important backed up on some old drives. I spent 2 weeks of what I can only describe as hell. I had so much data to scrub through and for tools to comb through that I would have to run them for a whole day, only to find that it failed at like 25%. And the problem is that losing 75% of data means I lost more like 25% because the corruption was all over the place randomly. I had wedding videos that had chunks of missing audio and video.

All you need is one serious data loss scare like that and you lose all trust in it.

If you're curious, I've been running those with XFS for a few years, and right now I'm testing bcachefs. No corruption issues, yet

[xinorom]

Not a troll. I've tried btrfs 3 times and had terrible experiences each time. The first time was a long time ago, like 2010 and my system lasted a few weeks before I had to reinstall. The second time was in 2013 and I had issues with file corruption

Btrfs was still marked as experimental back then. So you decided to use beta software for a production use case and now you want us to believe the current state of Btrfs is "untrustworthy" because you used it way before it was ready? I've been using it since about 2010 too and have never had a single issue with it, although in the early days I wouldn't have blamed anyone except myself if I had.

If you're curious, I've been running those with XFS for a few years, and right now I'm testing bcachefs. No corruption issues, yet

Those are both great filesystems, but one is not in the same league as Btrfs w.r.t. to features and the other seems further away from being mainlined than the author wants to believe. Also, you seem to be making the exact same mistake as you made before. I hope bcachefs doesn't eat your data and make you ragequit again.

[profoundWHALE]

Btrfs was still marked as experimental back then. So you decided to use beta software for a production use case and now you want us to believe the current state of Btrfs is "untrustworthy" because you used it way before it was ready? I've been using it since about 2010 too and have never had a single issue with it, although in the early days I wouldn't have blamed anyone except myself if I had.

That's why in 2010 I never put anything on it I cared about. I was testing it to see how it panned out.

I tried it again in 2013 because people called it 'stable'. Nope.

The one that really bothered me was from 2016. By that time, again, people had been saying btrfs is awesome and doesn't have any of the problems! Nope.

Those are both great filesystems, but one is not in the same league as Btrfs w.r.t. to features and the other seems further away from being mainlined than the author wants to believe. Also, you seem to be making the exact same mistake as you made before. I hope bcachefs doesn't eat your data and make you ragequit again.

Features are useless if it corrupts stuff and the scrubbing or fsck doesn't work or sometimes is the cause of the problem.

It's like having this super awesome railgun that you can shoot people with, but also sometimes it shoots backwards. Oops!

Maybe when trying to replace something like ZFS which is rock-solid stable, you might want to make sure your filesystem doesn't just eat your data.

Bcachefs hasn't eaten my data despite using the same drives in a similar configuration. The only headache that I can run into is due to it not being mainlined yet.

[xinorom]

I tried it again in 2013 because people called it 'stable'. Nope.

Which "people"? Pretty sure no one you should have been listening to was calling it stable in 2013.

fsck doesn't work or sometimes is the cause of the problem.

That sounds like a problem I've heard several times before, where people were just mashing keyboard and running random commands hoping it'd fix their system. There were explicit warnings about certain fsck options in the docs.

It's like having this super awesome railgun that you can shoot people with, but also sometimes it shoots backwards. Oops!

In your case it sounds more like giving an mildly retarded child a hand grenade and telling him not to pull the pin out, but knowing he definitely will anyway.

Maybe when trying to replace something like ZFS which is rock-solid stable, you might want to make sure your filesystem doesn't just eat your data.

Maybe when trying to replace a mature filesystem, you might want to actually use another mature filesystem instead of an early stage, beta filesystem that's been clearly labelled as such by it's developers?

Bcachefs hasn't eaten my data despite using the same drives in a similar configuration

I think you're in for a surprise if you think bcachefs is going to be plain sailing...

[profoundWHALE]

Which "people"? Pretty sure no one you should have been listening to was calling it stable in 2013.

Well now I know that you don't know what you're talking about. Maybe you should go troll somewhere else

https://git.kernel.org/pub/scm/linux...cb5c58097b918e

That sounds like a problem I've heard several times before, where people were just mashing keyboard and running random commands hoping it'd fix their system. There were explicit warnings about certain fsck options in the docs.

Exactly, fsck is known to break the filesystem. It's a complaint I listed because if you have a serious problem and fsck is more likely to destroy your data than to recover it, maybe it's poorly designed.

And no, I didn't just fsck it and type in whatever commands. I read up on the manuals and only did "safe" commands, the problem I had is they kept failing at like 75%.

Geez you need to learn to read.

In your case it sounds more like giving an mildly retarded child a hand grenade and telling him not to pull the pin out, but knowing he definitely will anyway.

You've got a serious case of projection my friend.

Maybe when trying to replace a mature filesystem, you might want to actually use another mature filesystem instead of an early stage, beta filesystem that's been clearly labelled as such by it's developers?

Again, you clearly don't know what you're talking about because it was "supposed" to be stable in 2013 and it sure as heck should have been stable by 2016.

But that's not really my complaint. If the filesystem fails, which happens, I should be able to run a file system check that doesn't destroy the filesystem.

I think you're in for a surprise if you think bcachefs is going to be plain sailing...

You're seriously retarded if you think that I don't know that. I build the package from source and install on root. I'm the one who posts the Mega download links for the Deb packages on Reddit.

But guess what? Every time I've had an issue, either the filesystem can fix itself, or Kent pushes an update that day.