Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
New filesystem: Btrfs! (cont.)
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3, ... 9, 10, 11  Next  
Reply to topic    Gentoo Forums Forum Index Gentoo Chat
View previous topic :: View next topic  
Author Message
ToeiRei
Veteran
Veteran


Joined: 03 Jan 2005
Posts: 1191
Location: Austria

PostPosted: Wed Jun 30, 2010 2:36 pm    Post subject: Reply with quote

I didn't give up yet as the system feels better with the latest patchset.
_________________
Please stand by - The mailer daemon is busy burning your messages in hell...
Back to top
View user's profile Send private message
Dont Panic
Guru
Guru


Joined: 20 Jun 2007
Posts: 322
Location: SouthEast U.S.A.

PostPosted: Wed Jun 30, 2010 5:12 pm    Post subject: Reply with quote

I haven't given up on Btrfs yet either.

But, on the other hand, I don't consider Btrfs anywhere near stable.

I don't want to be seen as flaming Btrfs, because I still think it has some really cool potential. But I think many users who are considering adopting Btrfs would be better served by a more honest appraisal of Btrfs' current stage of development.

I think people assume it's production ready just because it's in the stable kernel. But Btrfs is still under heavy development.

I remember how ext4 was getting flamed during the period it was calling itself "ext4-dev" in the kernel tree. Some people felt the transition from ext4-dev to ext4 was not smooth enough, and introduced some issues. But at least is was clear the ext4 was still in the developmental phase.

Btrfs is no-where near as stable as ext4 was when they dropped the "-dev" tag. There's a long list of issues to be addressed and features to be added that will continue to hold Btrfs under heavy development for the immediate future. Btrfs hasn't even finished writing it's fsck application yet.

I would like to see Btrfs be feature complete, and with far fewer issues rolling across the Mailing List before even trying to decide if it's ready for the "stable" label.
Back to top
View user's profile Send private message
ToeiRei
Veteran
Veteran


Joined: 03 Jan 2005
Posts: 1191
Location: Austria

PostPosted: Wed Jun 30, 2010 9:04 pm    Post subject: Reply with quote

Well... I'm using it here since quite a while on a single disk and didn't loose any data. For me it's quite stable and compression is a bliss as seen on portage. But I agree with keeping it away from the productive systems as there's no support for grub in example.

I'm curious about the next changes they'll do and how to migrate :)
_________________
Please stand by - The mailer daemon is busy burning your messages in hell...
Back to top
View user's profile Send private message
devsk
Advocate
Advocate


Joined: 24 Oct 2003
Posts: 2995
Location: Bay Area, CA

PostPosted: Wed Jun 30, 2010 10:19 pm    Post subject: Reply with quote

I am not flaming BTRFS. Just pointing out that it remains half-baked at this time. They need to cook it more before masses use it. Too early an adoption can backfire on BTRFS.
Back to top
View user's profile Send private message
Simba7
l33t
l33t


Joined: 22 Jan 2007
Posts: 706
Location: Billings, MT, USA

PostPosted: Thu Jul 01, 2010 2:22 am    Post subject: Reply with quote

ToeiRei wrote:
Well... I'm using it here since quite a while on a single disk and didn't loose any data. For me it's quite stable and compression is a bliss as seen on portage. But I agree with keeping it away from the productive systems as there's no support for grub in example.

I'm using it on 7 systems at the moment. Three remote servers, my router, my Portage server, my Cacti Server, and one other server. I also have it running on a PowerEdge 2950 at the local University. All the systems are using compression.

I'm still sticking with ext3 on the /boot partition. It seems to run flawlessly using that config. I run btrfs on everything else.
Back to top
View user's profile Send private message
Shining Arcanine
Veteran
Veteran


Joined: 24 Sep 2009
Posts: 1110

PostPosted: Thu Jul 01, 2010 3:35 am    Post subject: Reply with quote

Simba7 wrote:
ToeiRei wrote:
Well... I'm using it here since quite a while on a single disk and didn't loose any data. For me it's quite stable and compression is a bliss as seen on portage. But I agree with keeping it away from the productive systems as there's no support for grub in example.

I'm using it on 7 systems at the moment. Three remote servers, my router, my Portage server, my Cacti Server, and one other server. I also have it running on a PowerEdge 2950 at the local University. All the systems are using compression.

I'm still sticking with ext3 on the /boot partition. It seems to run flawlessly using that config. I run btrfs on everything else.


Btrfs seems to be either hit or miss for people.
Back to top
View user's profile Send private message
cach0rr0
Bodhisattva
Bodhisattva


Joined: 13 Nov 2008
Posts: 4123
Location: Houston, Republic of Texas

PostPosted: Thu Jul 01, 2010 3:40 am    Post subject: Reply with quote

Shining Arcanine wrote:

Btrfs seems to be either hit or miss for people.


pretty well been my experience. Brilliant on one laptop, terrible on the workstation/server where it sits atop dm-crypt vols.
Back to top
View user's profile Send private message
ToeiRei
Veteran
Veteran


Joined: 03 Jan 2005
Posts: 1191
Location: Austria

PostPosted: Thu Jul 01, 2010 4:46 pm    Post subject: Reply with quote

I agree. My PC is a laptop.
_________________
Please stand by - The mailer daemon is busy burning your messages in hell...
Back to top
View user's profile Send private message
Shining Arcanine
Veteran
Veteran


Joined: 24 Sep 2009
Posts: 1110

PostPosted: Thu Jul 01, 2010 5:10 pm    Post subject: Reply with quote

It was a disaster on my laptop with 2.6.33 RC 4.
Back to top
View user's profile Send private message
ToeiRei
Veteran
Veteran


Joined: 03 Jan 2005
Posts: 1191
Location: Austria

PostPosted: Thu Jul 01, 2010 5:19 pm    Post subject: Reply with quote

.34 or .35-rc3 had many bugfixes. So I can't complain
_________________
Please stand by - The mailer daemon is busy burning your messages in hell...
Back to top
View user's profile Send private message
Shining Arcanine
Veteran
Veteran


Joined: 24 Sep 2009
Posts: 1110

PostPosted: Thu Jul 01, 2010 7:56 pm    Post subject: Reply with quote

ToeiRei wrote:
.34 or .35-rc3 had many bugfixes. So I can't complain


I do not plan to use it until it is considered stable for production environments at the earliest, regardless of what success other people are having with it. I cannot afford to have my system's file system be rendered unusable by undeletable files or other kinds of weirdness.
Back to top
View user's profile Send private message
devsk
Advocate
Advocate


Joined: 24 Oct 2003
Posts: 2995
Location: Bay Area, CA

PostPosted: Thu Jul 01, 2010 8:12 pm    Post subject: Reply with quote

Shining Arcanine wrote:
ToeiRei wrote:
.34 or .35-rc3 had many bugfixes. So I can't complain


I do not plan to use it until it is considered stable for production environments at the earliest, regardless of what success other people are having with it. I cannot afford to have my system's file system be rendered unusable by undeletable files or other kinds of weirdness.
Major issue for me is lack of fsck. I suffered hidden corruptions twice and I could not get anything about it from fsck. No message about mismatch of checksums for those file when I tried to read them in 'dmesg'. What good are checksums if they are not used to verify and report on the integrity of data.
Back to top
View user's profile Send private message
ToeiRei
Veteran
Veteran


Joined: 03 Jan 2005
Posts: 1191
Location: Austria

PostPosted: Thu Jul 01, 2010 8:23 pm    Post subject: Reply with quote

I use to 'touch' the files and check against dmesg. That way I can see corruptions and act. Another good way is a tarball across the whole subvolume. (which are also good backups)
_________________
Please stand by - The mailer daemon is busy burning your messages in hell...
Back to top
View user's profile Send private message
dufeu
l33t
l33t


Joined: 30 Aug 2002
Posts: 924
Location: US-FL-EST

PostPosted: Fri Jul 09, 2010 6:47 pm    Post subject: Reply with quote

As has been noted here and elsewhere, btrfs is under heavy development.

I have a couple of thoughts I thought I'd throw out.
  1. Heavy development means just that. Anyone who sets up a completely btrfs based system is, in my book, a complete foole. Why? Because there are so many potential areas for problems to arise that it becomes impossible for any sane human to figure out what's going on. The best approach is to always install the simplest or near simplest case and go from there. I shake my head in disbelief at people who use btrfs in their '/' partition. Give me a break people! How do you figure out which problem (of many) you're looking at?

  2. When anything is under heavy development, being subscribed to the mailing list makes good sense. It doesn't matter if you understand less than one word in 10, there's still a lot of information to be gleaned from perusing the list. If you're following the thread here and thinking about installing btrfs or actually have an active btrfs based device, you should be subscribed to the mailing list too. Why? Not only can you get a sense of what's not working yet or what new features are currently being worked on, but you can get an idea of what's currently broken. This is valuable to know folks! If you have an active btrfs based partition and you're not subscribed to the list, I'm virtually rolling my eyes and shaking my head at you. Really.

  3. Always have a plan. Whenever you try something new, always ask yourself "What's the plan?" and then ask yourself "If it blows up, what's the fallback?" Why? You mean to tell me you're installing btrfs because it's "shiny?" {ooooh. Shiney! Must have/install!} {rolls eyes} {shakes head}

I'm taking the plunge with btrfs. Here's my 'plan'. The 'fallbacks' should be obvious. Note: each step in this list does have a fallback!
  1. Install btrfs on a single free partition on a non-critical drive. Rationale? Get practice with mkfs.btrfs. Get some practical understanding of some of the available options. {Done}

  2. Set up a simple files only server with multiple drives and a single instance of btrfs across those drives. Rationale? See how btrfs works across multiple devices in a relatively simple setting. {In Progress}

  3. Get more experience of what's involved. Write "How-To documentation". Why? Because writing down your experiences in the form of a How-to is an excellent tool for revealing shortfalls and/or misunderstandings of official documentation. {In Progress}

  4. Set up a more complex server. In addition to setting up a files serving partition based on btrfs, add things like '/var' and '/usr/portage/distfiles'. Rationale? The idea is to start stressing the btrfs service a bit more but still have a system you can recover if worst comes to worst. {Future}

  5. Set up a simple workstation for a non-techy user. Rational? This is where you finally should be trying out btrfs on '/' for the first time. There is a lot of wildy variant activity that happens at '/' as well as things like sockets, device nodes, etc. Far better to give a relatively lightweight user the first machine with '/' housed on a btrfs partition. It gives you a much better chance to see possible shortcomings. And it's easy to swap out a light weight system for something more stable if you need to. An understanding user who's co-operating with you will go a long way towards keeping stress out of your life. {Future}

  6. Set up a simple files only server with '/' on a btrfs based partition. Rationale? Assuming all the above has progressed reasonably well, now you can start taking a few, limited risks. {Future}

  7. And so on. Rationale? The idea is to break your risks down into manageable chunks. This is a critical concept. Never let your sense of "Shiny! Must have!" talk you into attempting to swallow chunks of risk you can't deal with.

OK. The above was a mouthful. This is what I've actually done/am doing:
  1. On my primary PC, I replaced the smallest drive (hd0) with a larger drive. The first 3 partitions on the drive are exact duplicates of the partitions of the drive that was replaced. The last partition was formated for btrfs and mounted as '/pub00'. This is where I tried out different mkfs.btrfs options etc. I did some initial copies of files to and from this partition. Note that no critical data or system files were involved at any time. This is where I also checked things like the functioning of nfs availability through /etc/exports and also determined what is involved in adding this partition to /etc/fstab. This is what linux reports about my primary PC hard drives:
    Code:
    brw-rw---- 1 root disk 8,  0 Jul  8 16:56 /dev/sda
    brw-rw---- 1 root disk 8,  1 Jul  8 20:56 /dev/sda1
    brw-rw---- 1 root disk 8,  2 Jul  8 16:56 /dev/sda2
    brw-rw---- 1 root disk 8,  3 Jul  8 20:56 /dev/sda3
    brw-rw---- 1 root disk 8,  4 Jul  8 16:56 /dev/sda4
    brw-rw---- 1 root disk 8, 16 Jul  8 16:56 /dev/sdb
    brw-rw---- 1 root disk 8, 17 Jul  8 20:56 /dev/sdb1
    brw-rw---- 1 root disk 8, 32 Jul  8 16:56 /dev/sdc
    brw-rw---- 1 root disk 8, 33 Jul  8 20:56 /dev/sdc1
    brw-rw---- 1 root disk 8, 48 Jul  8 16:56 /dev/sdd
    brw-rw---- 1 root disk 8, 49 Jul  8 20:56 /dev/sdd1
    brw-rw---- 1 root disk 8, 64 Jul  8 16:56 /dev/sde
    brw-rw---- 1 root disk 8, 65 Jul  8 20:56 /dev/sde1

    This is what btrfs-show reports:
    Code:
    Label: none  uuid: 079dc77d-7ac7-43c7-b282-e09964d25eea
            Total devices 1 FS bytes used 513.61GB
            devid    1 size 558.90GB used 558.90GB path /dev/sda4

    Btrfs v0.19-13-gb72e4c4-dirty

    This is what the /etc/fstab entry looks like:
    Code:
    /dev/sda4       /pub00          btrfs   noatime                 0 1

    I created a single device/partition btrfs using all default options. i.e.
    Code:
    mkfs.btrfs /dev/sda4

    If I've understood the available documentation correctly, this means that I've created a 'raid0' single device/partition btrfs with 2 copies of the meta data on /dev/sda4. Everything seems to work pretty much as expected. There are some observed "gotchas" I'll note below.

  2. I have a file server set up at my mother's house. It serves several important but non-critical functions. First, it serves as my personal backup for all the data I have on my personal workstation and laptops. I really do back up my systems and I really do have at least two copies of everything on at least two different systems. This includes all all the CDs and DVDs I've ripped from our personal collections as well as our collection of ebooks. {Thanks Project Gutenberg! Thanks Internet Archive!}

    Lately, we've had a number of power spikes at my mother's house due to lightning storms {she's located in Florida, US}. Lost a monitor, keyboard, mouse and attached electronic KVM switch. After looking over the server, I decided it was time to rebuild it. So I'm taking this opportunity to set up a simple btrfs based file server. The configuration is a total of 6 hard drives. The disk configuration looks like this:
    Code:
    brw-rw---- 1 root disk 8,  0 Jul  6 08:34 /dev/sda
    brw-rw---- 1 root disk 8, 16 Jul  6 08:34 /dev/sdb
    brw-rw---- 1 root disk 8, 32 Jul  6 08:34 /dev/sdc
    brw-rw---- 1 root disk 8, 48 Jul  6 08:34 /dev/sdd
    brw-rw---- 1 root disk 8, 64 Jul  6 08:34 /dev/sde
    brw-rw---- 1 root disk 8, 65 Jul  6 12:34 /dev/sde1
    brw-rw---- 1 root disk 8, 66 Jul  6 08:34 /dev/sde2
    brw-rw---- 1 root disk 8, 67 Jul  6 12:34 /dev/sde3
    brw-rw---- 1 root disk 8, 80 Jul  6 08:34 /dev/sdf
    brw-rw---- 1 root disk 8, 96 Jul  6 08:34 /dev/sdg

    This is what btrfs-show reports:
    Code:
    Label: PUBLIC  uuid: b71c7140-e845-4891-a8c9-98599be7d29c
            Total devices 5 FS bytes used 1.34TB
            devid    2 size 465.76GB used 372.00GB path /dev/sdb
            devid    5 size 233.76GB used 233.01GB path /dev/sdf
            devid    1 size 465.76GB used 371.02GB path /dev/sda
            devid    4 size 931.51GB used 372.26GB path /dev/sdd
            devid    3 size 931.51GB used 373.25GB path /dev/sdc

    Btrfs v0.19-13-gb72e4c4-dirty

    To create this server, I booted with 'System Rescue CD 1.5.6'. The sysrecuecd boots showing disks in traditional order. i.e. This PC has a PATA controller and 2 of the 6 disks are installed on it rather than all SATA. The /dev/sd# order of the boot disk is therefore different from the order I show here. I created the btrfs instance from the boot CD. So the command I show doesn't make sense for the results I display. This is the command I used to create the btrfs instance:
    Code:
    mkfs.btrfs -L PUBLIC /dev/sdb /dev/sdc /dev/sdd /dev/sde /dev/sdf
    where sda and sdb were the PATA drives and sdc, sdd, sde and sdf were the SATA drives.

    These are the relevant /etc/fstab entries:
    Code:
    /dev/sde1                       /boot           ext2    noauto,noatime          1 2
    /dev/sde2                       none            swap    sw,pri=1                0 0
    /dev/sde3                       /               ext4    noatime                 0 1
    /dev/disk/by-label/PUBLIC       /public         btrfs   noauto,noatime          0 1

    Note that '/', and by extension, '/usr', '/var', '/dev' and all other system directories/files are not on the btrfs based partition. I can't stress this enough. The idea is take a reasonable plunge in relatively calm water. The idea is not to jump off a pier while high tide is going out with the presence of powerful riptides.

    I'm still in the process of backing up files onto the server. Note that one of the hard drives is already full. I'll be checking the effectiveness of 're-balancing' later. :D

    I deliberately went with un-even sized drives because 1) I'm not buying any thing new and this is what I have and 2) I really want to see how well btrfs handles this situation since this is a design goal.


Now for a few 'gotchas' in no particular order:
  • You cannot create a btrfs instance and the go back to apply a label to it. i.e.
    Code:
    mkfs.btrfs /dev/sdb /dev/sdc /dev/sdd /dev/sde /dev/sdf
    mkfs.btrfs -L PUBLIC /dev/sdb

    What happens is that you end up with just /dev/sdb in the btrfs instance:
    Code:
    Label: PUBLIC  uuid: b71c7140-e845-4891-a8c9-98599be7d29c
            Total devices 1 FS bytes used 0.01GB
            devid    1 size 233.76GB used 0.01GB path /dev/sdb

    Once you see this in action, it makes sense. The progam is working as designed. But, this is an area I intend to explore further. In other file systems, applying a label to a disk has no effect on the partition instance. Since it's not clearly covered that the designed behavior is different from possible assumptions, you can really shoot yourself in the foot here.

  • In GUI based file managers, you'll find that you probably won't be able to write to most directories in a btrfs based instance without doing come CLI work with 'chmod' first. This is one of those areas under active development. This can result in all kinds of unexpected failures from programs expecting to write to or execute from a btrfs instance. You have been warned.

  • Apparently, by using raw /dev/sd# devices, the btrfs instance I show here won't automatically mount. Rather than go nuts with that, I turned off automount "-o noauto" in the /etc/fstab file. Instead, I added the following to /etc/conf.d/localmount:
    Code:
    btrfsctl -a &>/dev/null
    mount /public

    I'll be looking into this further. But not just yet.

I've written this long post so that you all can get an idea of the real world state of btrfs. It's works well enough for some real and interesting testing but not well enough for critical production installation. If you have non-critical PCs available to you, you can start like I'm doing to get a feel for it. So far, I will say I'm very pleased with what limited functional use I'm making of it. In the limited server environment I'm setting up, I think it will be a big improvement. But realize just how limited my installation is. There are only two users with no more than 8 computers able to pull data from it at any one time. This is not what you could call a 'mission critical' application. If I loose the server, it's not really a problem as I have copies of everything. So I get some valid real world testing at almost no risk.

Enjoy!
_________________
People whom think M$ is mediocre, don't know the half of it.
Back to top
View user's profile Send private message
gringo
Advocate
Advocate


Joined: 27 Apr 2003
Posts: 3793

PostPosted: Mon Aug 02, 2010 2:44 pm    Post subject: Reply with quote

i was quite happy with btrfs until a few days ago.
For some reason all the btrfs-* kernel daemons started to suck _a lot_ of cpu, rendering the box almost useless. I mean, just quiting htop could take 3 or 4 seconds f.ex. , firefox autocompletion could easily take 10 seconds, etc.
I tried to track down from where this could have come from ( initially i suspected it was firefox but it wasnt), upgrading the kernel made no difference either.

Now after reformating everything is normal again, did anyone have the same experience ?

cheers
Back to top
View user's profile Send private message
dufeu
l33t
l33t


Joined: 30 Aug 2002
Posts: 924
Location: US-FL-EST

PostPosted: Mon Aug 02, 2010 4:27 pm    Post subject: Reply with quote

gringo wrote:
i was quite happy with btrfs until a few days ago.
For some reason all the btrfs-* kernel daemons started to suck _a lot_ of cpu, rendering the box almost useless. I mean, just quiting htop could take 3 or 4 seconds f.ex. , firefox autocompletion could easily take 10 seconds, etc.
I tried to track down from where this could have come from ( initially i suspected it was firefox but it wasnt), upgrading the kernel made no difference either.

Now after reformating everything is normal again, did anyone have the same experience ?

cheers

Running btrfs on 3 different systems. On two systems, I'm running a multi device btrfs setup. On the third, I'm running btrfs on a single partition.

I am NOT running btrfs on my "root" partition on any of the three systems.

No - I haven't seen anything like that at this time.
_________________
People whom think M$ is mediocre, don't know the half of it.
Back to top
View user's profile Send private message
regomodo
Guru
Guru


Joined: 25 Mar 2008
Posts: 445

PostPosted: Tue Aug 03, 2010 5:40 pm    Post subject: Reply with quote

I've finally had enough of btrfs. My multi-device setup has broken for no reason and my patience has run out with it. Back to raid0 ext4.
Back to top
View user's profile Send private message
Letharion
Veteran
Veteran


Joined: 13 Jun 2005
Posts: 1342
Location: Sweden

PostPosted: Thu Aug 05, 2010 9:07 am    Post subject: Reply with quote

I have a strange bug with hardlinks, and I'm wondering if it's possibly related to btrfs.
I hope to attract some attention to https://bugs.gentoo.org/show_bug.cgi?id=329981 from someone with more btrfs knowledge than I have.
Back to top
View user's profile Send private message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 20048

PostPosted: Thu Aug 05, 2010 2:04 pm    Post subject: Reply with quote

gringo wrote:
Now after reformating everything is normal again
I know with ZFS, periodic scrubbing is recommended. Perhaps something similar with btrfs?
_________________
Quis separabit? Quo animo?
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6111
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Thu Aug 05, 2010 3:07 pm    Post subject: Reply with quote

pjp wrote:
gringo wrote:
Now after reformating everything is normal again
I know with ZFS, periodic scrubbing is recommended. Perhaps something similar with btrfs?


afaik online checking or even offline checking isn't fully possible yet for btrfs

meanwhile ZFS FTW ! ;)
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 20048

PostPosted: Thu Aug 05, 2010 7:46 pm    Post subject: Reply with quote

Maybe that is the actual problem. Deterioration over time. Since formatting "fixed" the problem.

I guess it depends on how long it took to reach that state, and whether or not a comprable time frame will lead to the same result.


EDIT: Split off from the original thread... seemed like a good demarcation.
_________________
Quis separabit? Quo animo?
Back to top
View user's profile Send private message
Simba7
l33t
l33t


Joined: 22 Jan 2007
Posts: 706
Location: Billings, MT, USA

PostPosted: Tue Aug 10, 2010 4:19 am    Post subject: Reply with quote

kernelOfTruth wrote:
meanwhile ZFS FTW ! ;)

I'd love to run it, but I'd have to switch a couple systems back to FreeBSD (which is tempting at times).
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6111
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Tue Aug 10, 2010 9:18 am    Post subject: Reply with quote

Simba7 wrote:
kernelOfTruth wrote:
meanwhile ZFS FTW ! ;)

I'd love to run it, but I'd have to switch a couple systems back to FreeBSD (which is tempting at times).


why ?

because I'm currently running it here with zfs-fuse :idea:
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
dufeu
l33t
l33t


Joined: 30 Aug 2002
Posts: 924
Location: US-FL-EST

PostPosted: Wed Aug 11, 2010 12:05 am    Post subject: Reply with quote

kernelOfTruth wrote:
pjp wrote:
gringo wrote:
Now after reformating everything is normal again
I know with ZFS, periodic scrubbing is recommended. Perhaps something similar with btrfs?


afaik online checking or even offline checking isn't fully possible yet for btrfs

The btrfs wiki documentation hasn't been updated yet. The defrag command for btrfs is:
Code:
btrfs filesystem defrag <file>|<dir> [<file>|<dir>...]

More information can be had from the man pages:
Code:
man btrfs

I haven't tried it yet but I expect to try it soon. I need to both defrag and re-balance on one of my experimental servers. Example:
Code:
# btrfs filesystem show
failed to read /dev/sr0
failed to read /dev/pktcdvd/pktcdvd0
failed to read /dev/pktcdvd/sr0
Label: none  uuid: 079dc77d-7ac7-43c7-b282-e09964d25eea
        Total devices 1 FS bytes used 523.05GB
        devid    1 size 558.90GB used 558.90GB path /dev/sda4

Btrfs v0.19-16-g075587c-dirty

# df
Filesystem           1K-blocks      Used Available Use% Mounted on
rootfs               380497160 305378364  55790596  85% /
/dev/root            380497160 305378364  55790596  85% /
rc-svcdir                 1024       136       888  14% /lib64/rc/init.d
udev                     10240       232     10008   3% /dev
shm                    4090624      1096   4089528   1% /dev/shm
/dev/sdb1            961432072 765411128 196020944  80% /home
/dev/sda4            586051200 549195176  36856024  94% /pub00    <=====
/dev/sdc1            961432072 957611844   1378328 100% /pub01
/dev/sdd1            961432072 903971612   8622460 100% /pub02
/dev/sde1            961432072 904381420   8212652 100% /pub03

# btrfs filesystem defragment /pub00

# btrfs filesystem balance /pub00


Important note! The version of btrfs-progs in portage is not yet the version which includes the new btrfs commands. Use btrfs-progs-9999 to get the latest unstable updates!
_________________
People whom think M$ is mediocre, don't know the half of it.
Back to top
View user's profile Send private message
Ant P.
Watchman
Watchman


Joined: 18 Apr 2009
Posts: 6920

PostPosted: Wed Aug 11, 2010 1:51 am    Post subject: Reply with quote

I got an SSD for my laptop today so I thought I'd take a risk and use a single big btrfs root partition (+ ext4 /boot).

Few short questions:
If it's mounted with -o=ssd already, will I gain anything at all by having laptop-mode enabled too?
How big's /usr/portage with compression? Also, is it smart enough to leave uncompressable things alone?
Can I do anything useful with subvolumes on a setup like this?
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Gentoo Chat All times are GMT
Goto page Previous  1, 2, 3, ... 9, 10, 11  Next
Page 2 of 11

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum