xfs
[Top] [All Lists]

RE: Buffalo LS-Q4.0 Raid 5 XFS errors

To: "'Dave Chinner'" <david@xxxxxxxxxxxxx>
Subject: RE: Buffalo LS-Q4.0 Raid 5 XFS errors
From: "Kirk Anderson" <kirkande@xxxxxxxxx>
Date: Thu, 29 Mar 2012 20:09:25 -0500
Cc: <xfs@xxxxxxxxxxx>
In-reply-to: <20120329235200.GE18323@dastard>
References: <004b01cd0d3f$c6d4fbc0$547ef340$@tx.rr.com> <20120329064054.GP5091@dastard> <002501cd0dbe$24be21c0$6e3a6540$@tx.rr.com> <20120329213100.GA18323@dastard> <005101cd0df9$67f61250$37e236f0$@tx.rr.com> <20120329230317.GD18323@dastard> <005501cd0e03$c1ccaf00$45660d00$@tx.rr.com> <20120329235200.GE18323@dastard>
Thread-index: AQJDN+KroJXgxoDTqYYE2XHNDqipKwJMKPzUAcmTQJ8B/Q7P3QHm4bXZAm6bNJICD+HnhwGd5VBBlSUB/vA=
I really appreciate your help to this point.  I do not have the xfs_irecover
command available.  Do you think there is an rpm for it that would be
compatible with the flavor of Linux this box is running? 

root@LS-QLF55:~# uname -a
Linux LS-QLF55 2.6.22.7 #395 Thu May 21 22:24:49 JST 2009 armv5tejl unknown

 If so, where may I find it?  Since I do not have a backup of this it sounds
like I have nothing to lose in trying the xfs_repair or should I hold out
for xfs_irecover?  Please let me know.  Thanks, Kirk

-----Original Message-----
From: Dave Chinner [mailto:david@xxxxxxxxxxxxx] 
Sent: Thursday, March 29, 2012 6:52 PM
To: Kirk Anderson
Cc: xfs@xxxxxxxxxxx
Subject: Re: Buffalo LS-Q4.0 Raid 5 XFS errors

On Thu, Mar 29, 2012 at 06:29:14PM -0500, Kirk Anderson wrote:
> These two matched.
> 
> root@LS-QLF55:~# dd if=/dev/sda6 bs=512 count=1 2> /dev/null | hexdump 
> -C > dump_sda6.txt root@LS-QLF55:~# dd if=/dev/md2 bs=512 count=1 2> 
> /dev/null | hexdump -C > dump_md2.txt root@LS-QLF55:~# diff 
> dump_sda6.txt dump_md2.txt root@LS-QLF55:~#
> 
> root@LS-QLF55:~# dd if=/dev/md2 bs=512 count=4 2> /dev/null | hexdump 
> -C
> 00000000  58 46 53 42 00 00 10 00  00 00 00 00 2b 4e 92 c0
> |XFSB........+N..|
> 00000010  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00
> |................|
> 00000020  1b 5a c5 ad cc e3 40 11  87 4d f5 8e 9b f8 37 c0
> |.Z....@..M....7.|
> 00000030  00 00 00 00 20 00 00 07  00 00 00 00 00 00 01 00  |....
> ...........|
> 00000040  00 00 00 00 00 00 01 01  00 00 00 00 00 00 01 02
> |................|
> 00000050  00 00 00 30 01 5a 74 a0  00 00 00 20 00 00 00 00  |...0.Zt....
> ....|
> 00000060  00 00 80 00 3d 84 10 00  01 00 00 10 00 00 00 00
> |....=...........|
> 00000070  00 00 00 00 00 00 00 00  0c 0c 08 04 19 00 00 19
> |................|
> 00000080  00 00 00 00 00 04 50 c0  00 00 00 00 00 00 02 ce
> |......P.........|
> 00000090  00 00 00 00 0b ba 45 ab  00 00 00 00 00 00 00 00
> |......E.........|
> 000000a0  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00
> |................|
> 000000b0  00 00 00 00 00 00 00 02  00 00 00 10 00 00 00 30
> |...............0|
> 000000c0  00 0c 10 00 00 00 10 00  00 00 00 00 00 00 00 00
> |................|
> 000000d0  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00
> |................|
> *
> 00000800

The AGF, AGI and AGFL have all been zeroed. Something has overwritten them.
Your filesysetm is likely to be toast.

> 
> 
> root@LS-QLF55:~# xfs_db -c "sb 0" -c p -c "agf 0" -c p -c "agi 0" -c p
> /dev/md2
> cache_node_purge: refcount was 1, not zero (node=0xb1698)
> xfs_db: cannot read root inode (22)

And that means the zeroing has extended well into the filesystem, and your
root directory has been lost. There's really not that much that reapir can
do for you at this point execpt make a mess. There is no AGI left to find
where in-use inodes might live to recover them, and the directory structure
cannot be used to find them, either, so I think the only thing you can do
now is start on disaster recovery.

Christoph had a patch to xfs_repair that allowed it to run xfs_irecover like
functionality - I don't think he ever posted it, so you might just have to
find the original xfs_irecover utility and make use of that to extract
whatever you can from the busted filesystem.

Other than that, I think that there's little we can do to help you recover
the filesystem intact at this point....

Cheers,

Dave.
--
Dave Chinner
david@xxxxxxxxxxxxx

<Prev in Thread] Current Thread [Next in Thread>