ich6r dmraid ntfs busy error

[David Parr]

Hi Heinz and co,

I'm having a problem with a windows system that recently crashed.
It has two 250GB WD SATA drives striped.

I've gotten dmraid to recognize and load the drives, and device-mapper
is managing the file, but mounting the NTFS filesystem consistantly gives device
already mounted error. The problem is a bad $MFT/$BITMAP. (See errors below)

Is there any way to recover this data, or is it hosed?

Thanks,
david




localhost mapper # dmraid -cs
isw_bidbgjihfe_RAID_Volume0

localhost mapper # dmraid -s -s -g
*** Superset
name   : isw_bidbgjihfe
size   : 976562496
stride : 0
type   : GROUP
status : ok
subsets: 1
devs   : 2
spares : 0
--> Active Subset
name   : isw_bidbgjihfe_RAID_Volume0
size   : 976561664
stride : 256
type   : stripe
status : ok
subsets: 0
devs   : 2
spares : 0

localhost mapper # dmsetup ls
isw_bidbgjihfe_RAID_Volume0     (253, 0)
isw_bidbgjihfe_RAID_Volume01    (253, 1)
localhost mapper # fdisk /dev/mapper/isw_bidbgjihfe_RAID_Volume0

The number of cylinders for this disk is set to 60788.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
    (e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): p

Disk /dev/mapper/isw_bidbgjihfe_RAID_Volume0: 499.9 GB, 499999571968 bytes
255 heads, 63 sectors/track, 60788 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

                                    Device Boot      Start         End 
Blocks   Id  System
/dev/mapper/isw_bidbgjihfe_RAID_Volume0p1   *           1       60787 
488271546    7  HPFS/NTFS

Command (m for help): q

localhost mapper # mount -t ntfs -o ro /dev/mapper/isw_bidbgjihfe_RAID_Volume0 
/mnt/windows/
mount: /dev/mapper/isw_bidbgjihfe_RAID_Volume0 already mounted or /mnt/windows/ busy

localhost mapper # uname -a
Linux localhost 2.6.11-gentoo-r4 #4 SMP Tue Mar 22 04:45:20 CST 2005 i686 
Intel(R) Pentium(R) 4 CPU 3.40GHz GenuineIntel GNU/Linux

localhost mapper # dmesg | grep -i ntfs
NTFS driver 2.1.22 [Flags: R/W DEBUG MODULE].
NTFS-fs error (device dm-1): ntfs_read_locked_attr_inode(): First extent of 
attribute has non-zero lowest_vcn.
NTFS-fs error (device dm-1): ntfs_read_locked_attr_inode(): Failed with error 
code -5 while reading attribute inode (mft_no 0x0, type 0xb0, name_len 0). 
Marking corrupt inode and base inode 0x0 as bad.  Run chkdsk.
NTFS-fs error (device dm-1): load_system_files(): Failed to load $MFT/$BITMAP 
attribute.
NTFS-fs error (device dm-1): ntfs_fill_super(): Failed to load system files.



-- 
David Parr                                   Phone: 512-232-2632
david.parr at mail.utexas.edu                   Office: MBB 3.232
Institute for Cellular and Molecular Biology http://www.icmb.utexas.edu
University of Texas at Austin                http://www.utexas.edu