xfs
[Top] [All Lists]

Re: partition recovery

To: Arkadiusz Miskiewicz <misiek@xxxxxxxxxx>
Subject: Re: partition recovery
From: Steve Lord <lord@xxxxxxx>
Date: 08 Nov 2001 10:22:09 -0600
Cc: linux-xfs@xxxxxxxxxxx
In-reply-to: <20011108163855.A19171@ikar.t17.ds.pwr.wroc.pl>
References: <20011108163855.A19171@ikar.t17.ds.pwr.wroc.pl>
Sender: owner-linux-xfs@xxxxxxxxxxx
On Thu, 2001-11-08 at 09:38, Arkadiusz Miskiewicz wrote:
> Hi,
> 
> Recently I've crashed both partition tables (1 and 2 copy).
> My setup was:
>       start           end
> hda1      0.031       10001.250       fat
>       10001.404       29313.918       extended lba
> hda5  10001.435       20002.807       xfs
> hda6  20002.838       20198.913       swap
> hda7  20198.944       20598.969       ext2
> hda8  20599.000       23603.312       xfs
> hda9  23603.344       29313.918       lvm
> (values in MB as shown by parted)
> 
> Now I'm trying to recover xfs partitions. Unfortunately
> I can't create exactly same partition with same positions
> as previously. Now fdisk, cfdisk, parted always create
> partition like this:
> 
> 2      10001.250  20002.992  primary
> (rounding to cylinder?)
> 
> So I'm trying to use xfs_repair (1.2.0) on such (not exacly
> same as in original) partition but. 
> 
> xfs_repair -n /dev/hda2
> Phase 1 - find and verify superblock...
> bad primary superblock - bad magic number !!!
> ......................................
> ...
> attempting to find secondary superblock...
> nable to verify superblock, continuing...
> ...
> 
> It found more than 5 backup superblocks but it was
> unable to verify all superblocks.
> 
> Any hints how to recover data from these partitions?
> 
> i686, 2.4.13, xfs-20011026
> -- 
> Arkadiusz MiÅkiewicz, AM2-6BONE    [ PLD GNU/Linux IPv6 ]
> http://www.t17.ds.pwr.wroc.pl/~misiek/ipv6/   [ enabled ]

You really need to get the partition back to exactly the same blocks
it used to be at - and that is your problem here, unless you have
partition table output which is in sector format, getting the same
ones back again is really hard. You could look at the raw disk for the
xfs superblock, this is placed in sector zero of the partition,
so it would be a good indication of the sector you need the partition
to start at.

The superblock starts with the string XFSB, so dumping the disk out
you would see something like this:

 od -c /dev/hda | more
0000000   Ã   Ã   |   l   b   a   L   I   L   O 001  \0 025 004   Z  \0
0000020  \0  \0 001   Ã   S 024   S   ; 030 002 200   y 001 031 002 200
0000040   y 001 027 002 200   y 001 001  \0  \0  \0  \0  \0  \0  \0 033
0000060 002 200   y 001  \a 001 200   G 001  \b 001 200   G 001  \t 001
0000100 200   G 001  \n 001 200   G 001  \v 001 200   G 001  \f 001 200
0000120   G 001  \r 001 200   G 001 016 001 200   G 001 017 001 200   G
0000140 001 020 001 200   G 001 021 001 200   G 001  \0  \0  \0  \0  \0
0000160  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0   Â
0000200   Ã  \a 216   Ã 214 006   z  \0 211   6   x  \0 211 036   |  \0
0000220 210 026   ~  \0   Â  \0 232 216   Ã   Â  \0 001   )   Ã   )   Ã
0000240   Ã   Ã   Â   Ã   Â  \0  \0 232 216   Ã   Â  \0 220 216   Ã   Â
0000260  \0   Â   Ã   Â  \r   Ã   i  \0   Â  \n   Ã   d  \0   Â   L   Ã
0000300   _  \0   Â   4  \0   h  \0  \v  \a   1   Ã   Â 221   Â   Â   `
0000320   u 017   N   Â 211   Ã  \t   Ã   t 034   Â   Â 002   Ã 023   Ã
0000340 016 210   Ã   Â   Ã   Ã       u 002   0   Ã 227   Ã   ;  \0   r
0000360 017 200   Ã 002   Ã   Ã   Â   I   Ã   &  \0   Ã  \0  \0  \0  \v
0000400   Â       Ã 034  \0   Ã 006  \0   1   Ã   Ã 023   Ã   Â   Ã   Ã
0000420 004   Ã 003  \0   Ã   Ã 004   $ 017 004   0   <   :   r 002 004
0000440  \a   P   0   Ã   Â 016   Ã 020   X   Ã   V   Q   S 210   Ã 200
0000460   Ã 217   Ã   Ã   @   u   3   Â   Â   U   Â  \0   A   Ã 023   r
0000500   ) 201   Ã   U   Â   u   #   Ã   Ã 001   t 036   [   Y 036   1
0000520   Ã   V   V   W   Q 006   S   j 001   j 020 211   Ã 026 037   Â
0000540  \0   B   Ã 023 215   d 020 037   Ã   D   [   Y   S   R   W   Q
0000560 006   Â  \b   Ã 023  \a   r   8   Q   Ã   Ã 006 206   Ã 211   Ã
0000600   Y 210   Ã   Ã   Ã 200   Ã   ?   Ã   Ã 226   X   Z   9   Ã   s
0000620   #   Ã   Ã   9   Ã   w 035   Ã   Ã 006 206   Ã 222   Ã   Ã   Ã
0000640   Ã  \0   Ã 211   Ã   Z   [ 206   Ã   Â 001 002   Ã 023   ^   Ã
0000660   Y   _   Ã 002   Â   @  \0  \0  \0  \0  \0  \0  \0  \0 200 001
0000700 001  \0 203   Ã   ?   A   ?  \0  \0  \0 203   - 020  \0  \0  \0
0000720 001   B 005   Ã   Ã   Ã   Ã   - 020  \0   Ã 002   ! 001  \0  \0
0000740  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0
0000760  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0   U   Â
0001000  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0
*
0077000   X   F   S   B  \0  \0 020  \0  \0  \0  \0  \0  \0 002 005   Â
0077020  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0
0077040   ~   Ã 034 024   z 004 021   Ã 222   Ã   Ã   B 200   $   Ã   Ã
0077060  \0  \0  \0  \0  \0 002  \0 004  \0  \0  \0  \0  \0  \0  \0 200
0077100  \0  \0  \0  \0  \0  \0  \0 201  \0  \0  \0  \0  \0  \0  \0 202
0077120  \0  \0  \0 020  \0  \0   @   Â  \0  \0  \0  \b  \0  \0  \0  \0
0077140  \0  \0 004   Â     204 002  \0 001  \0  \0 020   /   b   o   o
0077160   t  \0  \0  \0  \0  \0  \0  \0  \f  \t  \b 004 017  \0  \0 031
0077200  \0  \0  \0  \0  \0  \0  \0 200  \0  \0  \0  \0  \0  \0  \0   @
0077220  \0  \0  \0  \0  \0 001   Ã   Ã  \0  \0  \0  \0  \0  \0  \0  \0
0077240  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0
0077260  \0  \0  \0  \0  \0  \0  \0 002  \0  \0  \0  \0  \0  \0  \0  \0
0077300  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0
*
0100000   X   A   G   F  \0  \0  \0 001  \0  \0  \0  \0  \0  \0   @   Â
0100020  \0  \0  \0 001  \0  \0  \0 002  \0  \0  \0  \0  \0  \0  \0 001
0100040  \0  \0  \0 001  \0  \0  \0  \0  \0  \0  \0  \b  \0  \0  \0  \v
0100060  \0  \0  \0 004  \0  \0 023 233  \0  \0  \v   Ã  \0  \0  \0  \0
0100100  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0

Here you see the superblock, followed by the AGF structure.

Best of luck

   Steve


-- 

Steve Lord                                      voice: +1-651-683-3511
Principal Engineer, Filesystem Software         email: lord@xxxxxxx


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