Hello, I am having problems with my btrfs filesystem, I have a 1TB RAID-1 setup that I use as /home. When accessing certain programs, or moving large amounts of data, I often get errors looking like the one attached, after I get the error, /home becomes unusable and I reboot. I have tried running scrub, but it aborts (output given in the attachment). I am using the 3.2.0-rc4 kernel (as distributed in Debian) at the moment, but this may not have been the kernel I was using when I first started getting the errors. Thanks, Chris
On Tue, Dec 20, 2011 at 10:34:20PM +0000, Chris Baines wrote:> Hello, > > I am having problems with my btrfs filesystem, I have a 1TB RAID-1 > setup that I use as /home. When accessing certain programs, or moving > large amounts of data, I often get errors looking like the one > attached, after I get the error, /home becomes unusable and I reboot. > I have tried running scrub, but it aborts (output given in the > attachment). > > I am using the 3.2.0-rc4 kernel (as distributed in Debian) at the > moment, but this may not have been the kernel I was using when I first > started getting the errors.It looks like you have both virtual box and nvidia loaded here, so it is hard to tell exactly where this problem came from. Could you please grab the latest btrfs-progs and do a: btrfs inspect logical 714427781120 /home This will tell us a little more about the bad block. -chris -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 20 December 2011 23:36, Chris Mason <chris.mason@oracle.com> wrote:> On Tue, Dec 20, 2011 at 10:34:20PM +0000, Chris Baines wrote: >> Hello, >> >> I am having problems with my btrfs filesystem, I have a 1TB RAID-1 >> setup that I use as /home. When accessing certain programs, or moving >> large amounts of data, I often get errors looking like the one >> attached, after I get the error, /home becomes unusable and I reboot. >> I have tried running scrub, but it aborts (output given in the >> attachment). >> >> I am using the 3.2.0-rc4 kernel (as distributed in Debian) at the >> moment, but this may not have been the kernel I was using when I first >> started getting the errors. > > It looks like you have both virtual box and nvidia loaded here, so it is > hard to tell exactly where this problem came from. > > Could you please grab the latest btrfs-progs and do a: > > btrfs inspect logical 714427781120 /home > > This will tell us a little more about the bad block. > > -chris >How do I do this exactly, do I build a kernel from git? Thanks, Chris -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 12/21/2011 07:39 AM, Chris Baines wrote:> On 20 December 2011 23:36, Chris Mason <chris.mason@oracle.com> wrote: >> On Tue, Dec 20, 2011 at 10:34:20PM +0000, Chris Baines wrote: >>> Hello, >>> >>> I am having problems with my btrfs filesystem, I have a 1TB RAID-1 >>> setup that I use as /home. When accessing certain programs, or moving >>> large amounts of data, I often get errors looking like the one >>> attached, after I get the error, /home becomes unusable and I reboot. >>> I have tried running scrub, but it aborts (output given in the >>> attachment). >>> >>> I am using the 3.2.0-rc4 kernel (as distributed in Debian) at the >>> moment, but this may not have been the kernel I was using when I first >>> started getting the errors. >> It looks like you have both virtual box and nvidia loaded here, so it is >> hard to tell exactly where this problem came from. >> >> Could you please grab the latest btrfs-progs and do a: >> >> btrfs inspect logical 714427781120 /home >> >> This will tell us a little more about the bad block. >> >> -chris >> > > How do I do this exactly, do I build a kernel from git? >Only building the btrfs-progs will be ok. To get btrfs-progs: git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-progs.git thanks, liubo> Thanks, > > Chris > -- > To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html >-- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
I have built the btrfs-progs from git, and get this output when running the command: ioctl ret=-1, error: No such file or directory Thanks, Chris -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
I still haven''t managed to fix the problem/problems, and now I cant mount the partition. I have attached the portion from kern.log that I think is relevant. I also tried btrfsck, but it segfaulted, the log (btrfsck.log) is attached. When I try mounting now, the command doesn''t end. Any tips, Thanks, Chris
First thing I''d try is to try mounting it readonly ("mount ... -o
ro"). This should get it mounted, or at least failing at a later
point with different (better?) error messages to work from. (dmesg
output after such an attempt would be useful).
On the vanilla kernel front, Ubuntu has kernel debs without any extra
patches, which will/should work with debian. (There may be a similar
pure debian thing somewhere already, but anyways).
If you see anything about dkms building modules when installing,
you''ll have to disable those in order to prevent the nvidia and
virtualbox modules from being built/installed with vanilla kernel.
Debs available from
http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.2.1-precise/
And the relevant dkms lines if applicable (module names may also vary,
but dkms status will list the relevant ones assuming dkms is in fact
in use):
"dkms status"
"dkms uninstall -m nvidia -v <from status> -k <from status,
something
something 3.2.1>"
"dkms uninstall -m vboxhost -v <from status> -k <from status,
something something 3.2.1>"
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs"
in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Right, I can mount read only. Running a find command across the filesystem, yields lots of: [ 7775.228342] btrfs: corrupt leaf, bad key order: block=756502351872,root=1, slot=1 type messages in dmesg, running a scrub gives: scrub status for 7d65d152-1200-4043-981a-06757707f1ba scrub started at Wed Jan 18 22:29:30 2012 and was aborted after 5575 seconds total bytes scrubbed: 1.26TB with 0 errors And in dmesg: [13370.886533] btrfs bad tree block start 714438135808 714427781120 [13370.962864] btrfs bad tree block start 714438135808 714427781120 [13371.681333] btrfs bad tree block start 714438135808 714427781120 [13371.724529] btrfs bad tree block start 714438135808 714427781120 [13371.724690] btrfs bad tree block start 714438135808 714427781120 [13371.724969] btrfs bad tree block start 714438135808 714427781120 [13371.725099] btrfs bad tree block start 714438135808 714427781120 [13371.725243] btrfs bad tree block start 714438135808 714427781120 [13371.725412] btrfs bad tree block start 714438135808 714427781120 [13371.725714] btrfs bad tree block start 714438135808 714427781120 On 18 January 2012 20:48, cwillu <cwillu@cwillu.com> wrote:> First thing I''d try is to try mounting it readonly ("mount ... -o > ro"). This should get it mounted, or at least failing at a later > point with different (better?) error messages to work from. (dmesg > output after such an attempt would be useful). > > On the vanilla kernel front, Ubuntu has kernel debs without any extra > patches, which will/should work with debian. (There may be a similar > pure debian thing somewhere already, but anyways). > > If you see anything about dkms building modules when installing, > you''ll have to disable those in order to prevent the nvidia and > virtualbox modules from being built/installed with vanilla kernel. > > Debs available from > http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.2.1-precise/ > > And the relevant dkms lines if applicable (module names may also vary, > but dkms status will list the relevant ones assuming dkms is in fact > in use): > > "dkms status" > "dkms uninstall -m nvidia -v <from status> -k <from status, something > something 3.2.1>" > "dkms uninstall -m vboxhost -v <from status> -k <from status, > something something 3.2.1>"-- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html