UDS is over! And in the customary wrap-up I stood up and told the audience what the Foundations team have been discussing all week. One of the items is almost certainly going to get a little bit of publicity.
We are going to be doing the work to have btrfs as an installation option, and we have not ruled out making it the default.
I do stress the emphasis of that statement, a number of things would have to be true for us to take that decision:
- btrfs would need to not be marked “experimental” in the kernel config; we understand that this is planned for 2.6.35, which is the kernel version we are expecting to ship in Maverick.
- btrfs is not currently supported by GRUB2 (our boot loader) or the installer; these pieces would need to be finished before Feature Freeze.
- If that happens, we may make it the default for Alpha releases to gain testing; that testing must go smoothly.
- The btrfs upstream must be happy with the idea.
- We must be happy with the idea.
It’s a tough gauntlet, and it would only made with the knowledge that production servers and desktops can be run on Lucid as a fully supported version of Ubuntu at the same time. I’d give it a 1-in-5 chance.
I agree whole heartedly. I’ve used btrfs and think it’s been damn stable. Performance has been hit and miss, but the only way it will get better is by use. I’m not sure it will be ready for prime time in the time period of maverick. I’ve not been impressed with ext4 at all. As a fall back consider XFS? You loose key features that btrfs has but it’s still better the ext[234]!
I hope this actually happens. I think it is a good idea to prepare for 12.04! If you start early you can spend a long time working out issues with the “foundation” early.
btrfs have a lot of enterprise features, that would be very useful for desktops.
Will there be developed GUI tools to utilize these cool features?
And if so, which does Canonical think are suited for Ubuntu users?
XFS is not uniformly better than ext4. Some operations (like deletion of a large hierarchy) are slower.
I agree with Iwalton’s statements.
The snapshot feature sets the foundation for a number of features.
Rollback ala Fedora 13 would certainly be benificial.
Hey, as long as it gets huge amounts of testing and tweaking before the next LTS I think we’ll be doing all right. Sooner is nice, but of course this is a filesystem we’re talking about. I don’t feel too strongly about the default but having it as an option will be huge.
You will have big problem with 4. You have no expertise in btrfs design (I haven’t seen single btrfs commit by Ubuntu/Canonical developer), the same problem Ted Tso pointed you recently with ext4. I fear that by doing this kind of “favor” to btrfs you will only make it look bad. Just like with Pulseaudio – you pushed it to users before it was ready, like Lennart wrote few days ago.
Crazy in my opinion. Kernel crashes and data corruption come daily on the btrfs mailing list, and filesystem is something you really, really don’t want to mess with too much unless it’s pretty damn stable, which is definitely not the case with btrfs considering the time its been around.
Why the rush? It feels like using your users as beta testers.
Great news Scott btrfs is fairly sweet after looking at the benchmarks for ssd in particular.
@Iwalton: XFS really!?? People complain enough about EXT4 corrupting stuff
It doesnt support the same FS size as EXT4 either. So id hope that they wouldnt switch to that.
Pingback: Breaking News: BTRFS Might Be The Default File System In Ubuntu 10.10
If the few issues that I’ve run into with btrfs have been fixed, then I’m all for it. However, btrfs must be stable WRT multi-device filesystems and subvolumes before it can be put as default. In 2.6.32, I had a btrfs that was grown to 2 devices and mounting it would crash the kernel with about a 95% probability (I don’t understand why it wasn’t 100%, but I’m not complaining since I was able to reconsolidate back to a single volume after purging the data I only needed temporarily—after about 24 hours).
this is the best news i heard from the UDS !
Btrfs is going to move linux forward, specially Ubuntu
system rollbacks by default ! OMG
fedora is already working hard on it and you guys should team up.
http://www.phoronix.com/scan.php?page=article&item=fedora_13_btrfs&num=1
imagine no more accidental system breakage! less support requests, a system that has your back.
i think everybody has been waiting for this to hit linux
if you guys can implement it as easy as “Time-slider” gnome integration in OpenSolaris , than we have a Winner !
I agree with Lwalton,btrfs performance is hit and miss, and a lot harder to predict than ext3/4.
My suggestion is to leave it simply as an option for the next couple of releases to gather more data.
Making it default at this point is insane.
Pingback: Btrfs Could Be the Default File System In Ubuntu Meerkat | JetLib News
Canonical should hire some fs engineer to work full time in btrfs if they really want to default to btrfs anytime soon…because btrfs is not going to be ready for such task soon. Don’t get me wrong, I’m using it as my root fs and I haven’t had any problems, but so I did with ext4 in the past, and I saw launchpad bugs hitting data corruption (and ext4 isn’t a complete fs rewrite). Btrfs is too young, even ZFS has only started to be deployed on serious real world tasks recently. Btrfs has its own problems, like the hardlink problem that some users have already hit (which AFAIK needs a disk format change to be fixed), the “log mode COW” patches posted on Tuesday (which require, guess what, a disk format change). And the disk format is also being changed very slightly in 2.6.34 to support the default root subvolume patches. And then, there’s the small fact that the fsck tool right now doesn’t fixes anything, it’s read-only. It’s probably going to take at least a full year to start being considered a serious replacement for Ext4.
Pingback: Ubuntu 10.10 mungkin akan menggunakan BTRFS | FOSS Malaya
Yeah… I agree to have BTRFS as an option for 10.10. But I’m more happy if somebody out there can patch nautilus to have the ‘opensolaris time slider’.
Holy smokes, I thought you were joking when we talked last week. SWEET! HAHA.
BTRFS would be a fantastic addition to Ubuntu! ZFS is awesome, and Linux would do well to have something that’s comparable. There will be problems along the way, but in the end it would be more than worth it. XFS is also great, but doesn’t really compete with ZFS, and probably not BTRFS once it’s had a chance to mature. I hope all the stars align, so that we can start playing with BTRFS soon!
@lwalton: XFS is not robust against power failures and system crashes. It would not be a good default.
Do it! Ubuntu has got to lead the way. Bring it on Scott. Lots of work to get it to that stage yet, but I like what I’m hearing.
Upper commenters… you’re obviously doing it wrong.
I can’t agree with that. In my opinion btrfs is far from finished and I don’t think it’s a good idea to make it an available option (let alone the default one) in Maverick just because it’s not considered to be “experimental” by the kernel developers any more. In my opinion it still needs a lot of defelopers’ testing before being offered to the wide public in an “idiot proof” desktop distribution like Ubuntu.
And while you’re at it, please take the opportunity to fix GRUB2, which has really been a disappointment, not least its tendency to mess up multi-boots when performing an upgrade !…
Henri
Pingback: BTRFS nuovo file system per Ubuntu? Si al 20%
Superb to hear this result. I have been using BTRFS for >24 months on one of my fileservers with 1.5TB of store, mitigating risks with automated backups and snapshots.
Even with prerelease -rc kernels, I’ve only encountered a couple of benign bugs (which weren’t data corrupting) which were quickly fixed, so have had a good experience, and I’ve won a lot of performance and data safety (from checksumming) against unstable disks/SSDs.
It is a solid step forward to offer this as an install-time option, and certainly evaluating it from the first Maverick alpha will make good sense to judge stability; I’ll be contributing my testing to make this a great OS.
Will I still be able to use 10.10 MM on my ext2fs partition inherited from before 9.04 when i was using Debioan?
Or is there a way, non-destructive-ly, to go to maybe ext4fs in the meantime from that ext2fs partition?
Hello,
great news. I hope the ubuntu team can make this happen. Wide-scale testing is the only way forward for btrfs
Pingback: 대체로 무해함 » Blog Archive » 우분투 10.10 기본 파일시스템 – Btrfs
Well.. maybe Ubuntu should better wait for another distribution which is more focussed on current and up-to-date software to switch the default fs to btrfs. (e.g Fedora
) It’s meant to be a solution for non-techies, for end-user and so I think stability is more important than an ultra-modern superduper cool fs in accordance with its philosophy.
Pingback: Najciekawsze Wiadomości z Sieci » Archiwum bloga » Btrfs domyślnym systemem plików w Ubuntu?
I still wonder what’s the whole point in making a change on Ubuntus default file system. That’s what hasn’t yet went publik neiter on kernelnews, #ubuntu-dev, this blog noe anywhere lese.
Someone please thell the less privileged developers *why* (these decisions are made)/discussions are held since I think ext4 is a good filesystem and XFS seems to be no option on desktops.
Some articles about BRTFS on the web only mention some restrictions of ext4 in comparison to BRTFS. However this seems to be a bit vague to me. Finally I also noticed BRTFS is less performant at the moment.
Well now. That’s a twist. After the Ubuntu installer has it made impossible to use a btrfs partition for so long. (You know, btrfs is working quite well, even with an “experimental” label. You just have to do a lot of manual post-installation cleanup, if you have a small netbook hard disk and need fs compression.)
I would disagree with making it default. Because of Lucid’s insistence to make ext4 default, it wiped my hard drives with an unwanted format. Ubuntu has had a major problem lately of changing the “standard programs” for every release for awhile now.
How are we ever going to gain market share when we cannot be stable from one release to the next?
It might be more interesting to learn why the foundations team are considering BTRFS that what it needs to do to get there. I’ve been following development with interest, but my personal interest is fundamentally different to a distro’s interest!
what will the (average or experienced) Ubuntu user gain if the default filesystem for his root (/) partition is btrfs?
Do we have any plans like this one https://fedoraproject.org/wiki/Features/SystemRollbackWithBtrfs while installing updates?
Are there any boot performance gains or other sort of gains if we use btrfs?
Or we just want to default to btrfs just to live “on the cool egde” (aka “latest new and sexy filesystem” as mentioned in the blueprint summary https://wiki.ubuntu.com/FoundationsTeam/BtrfsSupport , https://blueprints.launchpad.net/ubuntu/+spec/foundations-lucid-btrfs-support)
Or we want it because fedora focuses on it too ?
system rollbacks and no more accidental breakage means for faster testing and developing.
even better experience for the novice user. almost foolproof
fedora and meego (btrfs by defauly) will be the first to benefit from this. Integration with Grub is the key.
Pingback: Btrfs potrebbe essere il nuovo File System di Ubuntu 10.10 Maverick
The claim that Ubuntu can’t offer btrfs as an option because “it’s supposed to be a n00b distro” is ridiculous. First, it’s used by all sorts of different people, including experienced Linux professionals. Second, that’s why there’s a distinction between an option and the default – inexperienced users shouldn’t be messing with settings away from the defaults, but people who know why they would want to and what they mean should be able to.
Test it. Then test it again. Then test it again. Then and only then start talking about making it the default. Release it when it’s ready and you have full confidence in it, not when some deadline arrived. But start the testing now, don’t postpone it just because the decision to put it in hasn’t been made yet.
@Shane Fagan – ” doesnt support the same FS size as EXT4 either.” Heh, true, xfs supports well beyond 16T, ext4 doesn’t yet.
@alex, xfs is quite capable of dealing with power loss etc, but like any journaling filesystem, it needs to have properly configured storage w.r.t. drive caches, write barriers, etc.
But more on topic, I have to wonder how on earth Canonical expects to support enterprise customers on BTRFS without any developer expertise or commitment.
Pingback: Najciekawsze Wiadomości z Sieci » Archiwum bloga » Niewielkie szanse na Btrfs w Ubuntu 10.10
Pingback: Ubuntu 10.10 Mungkin Guna Btrfs Ganti EXT4- syafiq.net
Pingback: UDS-M Day 5 « Working with Ubuntu
I don’t agree that XFS is not good for the desktop. I’ve run it as my primary file system for nearly a decade. Performance is usually on par or better in my day to day use, and I’ve never lost data. 16 exabyte volumes, and 8 exabyte file sizes? those are large numbers, can’t imagine exceeding them in the near future.
On the server XFS is a awesome choice period.
I’ve been exploring BTRFS on both servers and workstations and have been pretty happy.
The best reasons to examine both are last:
BTRFS and XFS have great user land tools that make them both slam dunks.
With XFS being able to to make snapshots, online defrags and re-sizing, real time volumes, Extent based allocations, and on and on.
BTRFS, has remarkable feature set also, some of the above plus ones at are novel to Linux.
there are new features discussed, one thats been talked about is deduplication, now that would be awesome.
Oh and one last thing, Ted Tso was right, Ubuntu needs a senior Kernel developer. And right now I can’t think of a better on then Christoph Hellwig. This guy knows his stuff! Mark Lord would be another good choice!
Pingback: Btrfs ar putea fi sistemul de fişiere implicit în Ubuntu 10.10 « Blogul lui ARVAT
Mabe it’s more impressive if the plan did not involve
“There’s a patch in legacy grub that looks like it could be coerced into grub2 or perhaps support will land in grub2 in the next few months and we pick this up for free.”
https://blueprints.edge.launchpad.net/ubuntu/+spec/foundations-m-btrfs-support
Pingback: Einblicke in Ubuntu 10.10 Maverick Meerkat | Ubuntuxx
In response to the anonymous author two comments above: the phrase you quote was a somewhat inaccurate transcription of a face-to-face conversation which didn’t reflect the conclusions we arrived at. In the process of drafting that specification properly, I rewrote it to reflect reality. We have in fact assigned a developer to add btrfs support to GRUB 2 – I’ve already done a fair bit of upstream work on GRUB 2 and mentioned this assignment to the current primary maintainer on IRC, so he’s aware of it.
The background here is that a Red Hat employee posted a fairly complete-looking patch for btrfs support against GRUB Legacy to grub-devel. It’s great that they sent this, but unfortunately GRUB Legacy no longer has an upstream to speak of and so the GRUB 2 maintainers asked if it could be ported to GRUB 2 instead. I never saw a public response to that (and of course RH haven’t switched to GRUB 2 yet, so it may not be too high on their priority list) – there may have been something privately that never reached the list. Certainly, nobody has yet done this and we expect that if we want it to happen in GRUB 2 then we’ll need to do it ourselves. Things are complicated somewhat because the patch includes a copy of btrfs.h, which is GPLv2-only while GRUB is GPLv3; GRUB upstream has asked for an exception to be made, but I gather at second hand that this is tied up in Oracle’s legal department. It’s unfortunately possible that we will need to clean-room this in order to comply with everyone’s licences, although the ideal situation would be that we could simply port the already-working patch.
I’ve already posted a patch for upstream review which at least fixes up grub-probe to cope with the way btrfs returns a virtual device number in st_dev, so that you can run btrfs / with a /boot that GRUB supports. I expect to be able to get this reviewed pretty soon, but I would hope we’d get rather further than that.
And, for what it’s worth, while we’ve been somewhat flippantly talking about using btrfs by default for 10.10, the odds are quite heavily against that being the right thing to do. Nevertheless, it offers some interesting possibilities and we’d like to explore them sooner rather than later, so there’s no harm in setting high goals so that we have the incentive to get the remaining bits of our userspace sorted out. I assume Red Hat ported GRUB Legacy to btrfs for similar reasons.
I’m happy Ubuntu is exploring other file system choices.
Both file sizes and volume sizes will be pressing issues sooner than later.
It seems crazy that I’d ever consider 16TB for files and 1EB a legitimate limitation (as in ext4). But I can see it quickly becoming a issue.
I’d never thought much about de-fragmentation, till I had file systems with hundred of thousands of files, many larger than two gigs.
Or that I would have volumes I needed to resize, with the expectation of *NO* downtime.
Or even having a file server that by default has eight or sixteen cores, And that I could/should/need to leverage them, to improve performance.
SUSE Linux Is Hooking Up With Btrfs Too
http://www.phoronix.com/scan.php?page=news_item&px=ODI2Ng
Pingback: BTRFS and Ubuntu « The-Source.com
Pingback: Sobre Btrfs « Doculinux
Pingback: Nano Theme » Btrfs May Be The Default File-System In Ubuntu 10.10 » Nano Theme
Pingback: Conociendo Btrfs » Doculinux