Help needed! ZFS I/O error recovery?

[Pawel Jakub Dawidek]

On Sun, Oct 04, 2009 at 04:22:30PM -0500, Aaron Hurt wrote:
> schroeder# zdb -ll /dev/ad4s1
[...]
>    txg=478240
[...]
> schroeder# zdb -ll /dev/ad6s1
[...]
>    txg=478204
[...]
> schroeder# zdb -ll /dev/ad8s1
[...]
>    txg=478240
[...]
> schroeder# zdb -ll /dev/ad10s1
[...]
>    txg=478240

As you can see one of your vdevs (ad6s1) has lower transaction group
number than the others. The difference is quite big (36 uberblocks), so
we may not be able to go that far in the past. It might be that ZFS
doesn't want to use ad6s1, because it is not up-to-date and there are
some errors on one of the other slices.

If you have you data backed up you may try this patch:

	http://people.freebsd.org/~pjd/patches/vdev_label.c.patch

Once you run ZFS with the patch you can try setting sysctl
vfs.zfs.maxtxg to 478204 and try importing your pool. It will try to
import the pool on txg from ad6s1. It won't work if the other slices
don't have this uberlock anymore or some earlier data is already
overwritten. Do it on your own risk, as it might mess up your data even
further.

-- 
Pawel Jakub Dawidek                       http://www.wheel.pl
pjd at FreeBSD.org                           http://www.FreeBSD.org
FreeBSD committer                         Am I Evil? Yes, I Am!
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-fs/attachments/20091004/3b78987e/attachment.pgp