Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 05 May 2011 13:04:59 +0200
From:      Stefan Esser <se@FreeBSD.org>
To:        FreeBSD Hackers <freebsd-hackers@freebsd.org>
Subject:   ZFS pool lost
Message-ID:  <4DC2845B.5040401@FreeBSD.org>
In-Reply-To: <4DC26533.5010904@FreeBSD.org>
References:  <4DC26533.5010904@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
Sorry for the follow-up to my own posting, but I noticed, that I left
out significant
information.

The system is an Athlon64 (X2, but was running with SMP disabled at the
time) in
32 bit mode (i386) with 4GB RAM running 8-STABLE in a ZFS only configuration
(one UFS partition was mounted but not accessed; it contains an
emergency system
which allows me to boot the exact kernel and access /rescue in exactly
the version that
was installed on ZFS). The panic occurs under 8-STABLE/i386 and with
identical
message when booting a -CURRENT live file-system from CD.

The ZPOOL consists of 3 1TB drives in a raidz1 configuration, of which
only some 40GB
were free.

The panic message reads (under -CURRENT, but it is the same when booting
8-STABLE):

panic: solaris assert: ss == NULL, file:
/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/space_map.c, line: 109

No dumps are currently available (I installed a -CURRENT into the
normally used SWAP
space), but I could try to add a new drive.

I'm willing to analyze and debug this, if there is any chance of
recovery. As I understand
the situation, the inconsistency in the ZPOOL has been caused by some
locking or
processing error during the file move and there is no chance, to
identify the cause of the
file-system corruption in this state.

Any ideas?

Best regards, STefan

Am 05.05.2011 10:52, schrieb Stefan Esser:
> I just lost my 2TB ZFS pool while moving some large files (ISO images)
> under a current stable. The system paniced with a message very similar
> to the one reported in
> 
> 	http://unix.derkeiler.com/Mailing-Lists/FreeBSD/current/2009-05/msg00181.html .
> 
> Seems, that the on-disk structure is damaged, since the identical panic
> happens even when I try to import into a fresh -CURRENT with its newer
> ZFS code.
> 
> Is there any chance to recover my data? I have copies of some critical
> information, but no easily restorable complete backups and thus I´m very
> interested in getting (most of) this pool working again ...
> 
> Any ideas or help?
> 
> Thanks in advance!
> 
> STefan




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4DC2845B.5040401>