Help repairing a corrupted superblock in an LVM

linux-lvm Chip Tondreau 1 participant 2 messages Nov 16, 2005

Share & tools

Chip Tondreau

author

2005-11-16 06:53 UTC #0 lore raw

Hi Folks,

I've been a bad newbie. After installing a new SuSE 10 installation using LVM
(4 drives, ~370GB "root" lv mounted at "/" and a 2GB "swap" lv), I migrated a
shedload of very important data to the server, I decided to add a fifth physical
drive to the volume. It turns out the drive was bad and on reboot, the LVM would not mount and the system would not start.

I purchased a new 300GB drive, unhooked the LVM drives and installed a new system on the new drive. Once this was completed, I reattached the LVM drives
(except the bad drive) and booted. SuSE successfully recognized the LVM and I was able to remove the bad drive definition using pvdelete. Here is the
current config:

server:/ # pvscan -v >
Wiping cache of LVM-capable devices
Wiping internal VG cache
Walking through all physical volumes
PV /dev/hda2 VG system lvm2 [74.29 GB / 40.00 MB free]
PV /dev/hdb1 VG system lvm2 [76.32 GB / 0 free]
PV /dev/hde1 VG system lvm2 [149.04 GB / 0 free]
PV /dev/hdg1 VG system lvm2 [74.50 GB / 0 free]
Total: 4 [374.14 GB] / in use: 4 [374.14 GB] / in no VG: 0 [0 ]

server:/ # lvscan -v
Finding all logical volumes
ACTIVE '/dev/system/swap' [2.00 GB] inherit
ACTIVE '/dev/system/root' [372.10 GB] inherit

server:/ # vgscan -v
Wiping cache of LVM-capable devices
Wiping internal VG cache
Finding all volume groups
Finding volume group "system"
Reading all physical volumes. This may take a while... Found volume group "system" using metadata type lvm2

Unfortunately, my attempt to mount the LVM fails because the superblock for the file system is not valid. It turns out that the number of blocks in the

superblock includes the blocks for the drive that was removed. dd and fsck return:

     server:/ # dd if=/dev/system/root of=/dev/null bs=1k count=1024
     1024+0 records in
     1024+0 records out
     1048576 bytes (1.0 MB) copied, 0.032456 seconds, 32.3 MB/s

     server:/ # fsck /dev/system/root
     Do you want to run this program?[N/Yes] (note need to type Yes if you

do):Yes
bread: Cannot read the block (127559679): (Invalid argument).

reiserfs_open: Your partition is not big enough to contain the filesystem of (127559679) blocks as was specified in the found super
block.

Failed to open the filesystem.

If the partition table has not been changed, and the partition is valid and it really contains a reiserfs partition, then the superblock is corrupted and you need to run this utility with --rebuild-sb.

Warning... fsck.reiserfs for device /dev/system/root exited with signal
6.

     fsck.reiserfs /dev/system/root failed (status 0x8). Run manually!

I have now reached the end of what little I know about this process. Does anyone
have any suggestions as to how to repair the superblock? I tried the --rebuild-sb
but it does not work (returns an "unknown -e parameter" error... Strange...)

Thanks,

Chip

Chip Tondreau

author

2005-11-16 13:58 UTC #1 lore raw

Never mind. It would appear that using reiserfsck rather than fsck is what is required.

   reiserfsck --rebuild-sb /dev/system/root
   reiserfsck --check  /dev/system/root
   reiserfsck --rebuild-tree  /dev/system/root

seems to have cured the problem. The file system is now mountable and short of 30 some odd files that were written to the bad drive, all the data is in tact!

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Chip Tondreau
Sent: Wednesday, November 16, 2005 1:53 AM
To: [email protected]
Subject: [linux-lvm] Help repairing a corrupted superblock in an LVM

Hi Folks,

I've been a bad newbie. After installing a new SuSE 10 installation using LVM
(4 drives, ~370GB "root" lv mounted at "/" and a 2GB "swap" lv), I migrated a shedload of very important data to the server, I decided to add a fifth physical drive to the volume. It turns out the drive was bad and on reboot, the LVM would not mount and the system would not start.

I purchased a new 300GB drive, unhooked the LVM drives and installed a new system on the new drive. Once this was completed, I reattached the LVM drives (except the bad drive) and booted. SuSE successfully recognized the LVM and I was able to remove the bad drive definition using pvdelete. Here is the current config:

server:/ # lvscan -v
Finding all logical volumes
ACTIVE '/dev/system/swap' [2.00 GB] inherit
ACTIVE '/dev/system/root' [372.10 GB] inherit

Unfortunately, my attempt to mount the LVM fails because the superblock for the file system is not valid. It turns out that the number of blocks in the

superblock includes the blocks for the drive that was removed. dd and fsck return:

     server:/ # dd if=/dev/system/root of=/dev/null bs=1k count=1024
     1024+0 records in
     1024+0 records out
     1048576 bytes (1.0 MB) copied, 0.032456 seconds, 32.3 MB/s

     server:/ # fsck /dev/system/root
     Do you want to run this program?[N/Yes] (note need to type Yes if you

do):Yes
bread: Cannot read the block (127559679): (Invalid argument).

reiserfs_open: Your partition is not big enough to contain the filesystem of (127559679) blocks as was specified in the found super
block.

Failed to open the filesystem.

Warning... fsck.reiserfs for device /dev/system/root exited with signal
6.

     fsck.reiserfs /dev/system/root failed (status 0x8). Run manually!

I have now reached the end of what little I know about this process. Does anyone have any suggestions as to how to repair the superblock? I tried the --rebuild-sb but it does not work (returns an "unknown -e parameter" error... Strange...)

Thanks,

Chip

_______________________________________________
linux-lvm mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/linux-lvm
read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/