unable to mount disk?

unable to mount disk?

[Matt Zagrabelny]

Greetings btrfs folks,

I am attempting to mount a filesystem on a (likely) failing disk:

$ mount /var/local/media
mount: /var/local/media: wrong fs type, bad option, bad superblock on
/dev/sdb, missing codepage or helper program, or other error.

excerpt from dmesg:

[  352.642893] BTRFS info (device sdb): disk space caching is enabled
[  352.642897] BTRFS info (device sdb): has skinny extents
[  352.645562] BTRFS error (device sdb): devid 2 uuid
bf09cc8f-44d6-412e-bc42-214ff122584a is missing
[  352.645570] BTRFS error (device sdb): failed to read the system array: -2
[  352.646770] BTRFS error (device sdb): open_ctree failed

Is there anything I can do to attempt to recover this hardware issue?

Thanks,

-m

Re: unable to mount disk?

[Chris Murphy]

On Tue, Sep 28, 2021 at 2:53 PM Matt Zagrabelny <mzagrabe@d.umn.edu> wrote:
>
> Greetings btrfs folks,
>
> I am attempting to mount a filesystem on a (likely) failing disk:
>
> $ mount /var/local/media
> mount: /var/local/media: wrong fs type, bad option, bad superblock on
> /dev/sdb, missing codepage or helper program, or other error.
>
> excerpt from dmesg:
>
> [  352.642893] BTRFS info (device sdb): disk space caching is enabled
> [  352.642897] BTRFS info (device sdb): has skinny extents
> [  352.645562] BTRFS error (device sdb): devid 2 uuid
> bf09cc8f-44d6-412e-bc42-214ff122584a is missing
> [  352.645570] BTRFS error (device sdb): failed to read the system array: -2
> [  352.646770] BTRFS error (device sdb): open_ctree failed
>
> Is there anything I can do to attempt to recover this hardware issue?


This is a 2+ device Btrfs file system? And one of the devices, devid2,
is missing. So you need to figure out why. If the drive has in fact
failed, rather than just missing power/data connection, you can mount
degraded using the 'degraded' mount option. Just be advised that there
are potentially some consequences to operating under degraded mode,
depending on the specific configuration.

-- 
Chris Murphy

Re: unable to mount disk?

[Matt Zagrabelny]

On Tue, Sep 28, 2021 at 1:59 PM Chris Murphy <lists@colorremedies.com> wrote:
>
> On Tue, Sep 28, 2021 at 2:53 PM Matt Zagrabelny <mzagrabe@d.umn.edu> wrote:
> >
> > Greetings btrfs folks,
> >
> > I am attempting to mount a filesystem on a (likely) failing disk:
> >
> > $ mount /var/local/media
> > mount: /var/local/media: wrong fs type, bad option, bad superblock on
> > /dev/sdb, missing codepage or helper program, or other error.
> >
> > excerpt from dmesg:
> >
> > [  352.642893] BTRFS info (device sdb): disk space caching is enabled
> > [  352.642897] BTRFS info (device sdb): has skinny extents
> > [  352.645562] BTRFS error (device sdb): devid 2 uuid
> > bf09cc8f-44d6-412e-bc42-214ff122584a is missing
> > [  352.645570] BTRFS error (device sdb): failed to read the system array: -2
> > [  352.646770] BTRFS error (device sdb): open_ctree failed
> >
> > Is there anything I can do to attempt to recover this hardware issue?
>
>
> This is a 2+ device Btrfs file system? And one of the devices, devid2,
> is missing. So you need to figure out why. If the drive has in fact
> failed, rather than just missing power/data connection, you can mount
> degraded using the 'degraded' mount option. Just be advised that there
> are potentially some consequences to operating under degraded mode,
> depending on the specific configuration.

Thanks for the hints, Chris. Appreciated.

What sort of potential consequences are we talking about, if you could
elaborate?

Cheers,

-m

Re: unable to mount disk?

[Chris Murphy]

On Tue, Sep 28, 2021 at 3:23 PM Matt Zagrabelny <mzagrabe@d.umn.edu> wrote:
>
> On Tue, Sep 28, 2021 at 1:59 PM Chris Murphy <lists@colorremedies.com> wrote:
> >
> > On Tue, Sep 28, 2021 at 2:53 PM Matt Zagrabelny <mzagrabe@d.umn.edu> wrote:
> > >
> > > Greetings btrfs folks,
> > >
> > > I am attempting to mount a filesystem on a (likely) failing disk:
> > >
> > > $ mount /var/local/media
> > > mount: /var/local/media: wrong fs type, bad option, bad superblock on
> > > /dev/sdb, missing codepage or helper program, or other error.
> > >
> > > excerpt from dmesg:
> > >
> > > [  352.642893] BTRFS info (device sdb): disk space caching is enabled
> > > [  352.642897] BTRFS info (device sdb): has skinny extents
> > > [  352.645562] BTRFS error (device sdb): devid 2 uuid
> > > bf09cc8f-44d6-412e-bc42-214ff122584a is missing
> > > [  352.645570] BTRFS error (device sdb): failed to read the system array: -2
> > > [  352.646770] BTRFS error (device sdb): open_ctree failed
> > >
> > > Is there anything I can do to attempt to recover this hardware issue?
> >
> >
> > This is a 2+ device Btrfs file system? And one of the devices, devid2,
> > is missing. So you need to figure out why. If the drive has in fact
> > failed, rather than just missing power/data connection, you can mount
> > degraded using the 'degraded' mount option. Just be advised that there
> > are potentially some consequences to operating under degraded mode,
> > depending on the specific configuration.
>
> Thanks for the hints, Chris. Appreciated.
>
> What sort of potential consequences are we talking about, if you could
> elaborate?
>
> Cheers,
>
> -m


If it's a 2 disk raid1, and one device dies, and you mount -o degraded
- it's no longer possible for btrfs to create raid1 block groups,
because that profile requires creating a mirrored pair of chunks which
can't happen because there's only one device. It's possible some data
will write to existing raid1 chunks, but other data can cause new
single chunks to be created. The data in these block groups won't
automatically be replicated when you do 'btrfs replace' to replace the
failed device, or if you fix the device problem and it rejoins the
working drive, a scrub won't catch up the stale drive either. You will
have to do 'btrfs balance start -dconvert=raid1,soft
-mconvert=raid1,soft" to make sure the single profile block groups are
converted to raid1 block groups.


-- 
Chris Murphy