Hi everyone,

I need help about my below query, i am using a linux (centos5) system on Vmware for operating my website, but 2 days ago, due to an electricity problem,i found my closed and i cannot open anymore my linux system and i am getting the following error code;


EXT3-fs error (device dm-0): ext3_check_descriptors: Block bitmap for group 384 not in group (block 1786349) !
EXT3-fs: group descriptors corrupted!
mount: error mounting /dev/root on /sysroot as ext3: Invalid argument
setuproot: moving /dev failed: No such file or directory
setuproot: error mounting /proc: No such file or directory
setuproot: error mounting /sys: No such file or directory
switchroot: mount failed: No such file or directory
Kernel panic - not syncing: Attempted to kill init!


Actually i am checked too many forums but i think that i can find the best solution over here and i hope to fix it soon..

Configure VMware to start your guest booting from a Centos installer or Live CD so you can work on the disk. If the disk contains one or more LVMs I suggest you check those first for consistency. And please note that while an issue may be urgent for you that is no reason to crosspost your thread across LQ fora.

Actually i want and trying to close this 2 posts with includes "urgent" at subjects and i re-opened a new post but i can't do that.. Anyway sorry for the confusion and revert to my above post

From your actions I understand you want to comply with the LQ Rules. But editing out the contents of your original post is not a good practice. It is also unnecessary because a moderator can easily help you by for instance changing the thread title. BTW you haven't responded to what I wrote.

Actually i am reverted to you at my other post but if you want we can continue it from here too.
Could any moderator can change the subject line with
How to fix "Group Descriptors Corrupted" Issue
Thanks..

Thanks for your helps, how can i check the consistency of my disk? i checked some posts about fsck command and i tried some of them but my lvm disk is still not working..
Could you advise me a way to check and fix it correctly?

We should find out the state of things first. Once you have booted with the installer or Live CD you should be able to run '(fdisk -l 2>&1; pvscan -d -v 2>&1; vgscan -d -v 2>&1)|tee -a /tmp/logfile', save /tmp/logfile outside your VM guest and post the contents here.

This is the outcome of log file as requested


Disk /dev/dm-0 doesn't contain a valid partition table
Disk /dev/dm-1 doesn't contain a valid partition table

Disk /dev/dm-0: 4294 MB, 4294967296 bytes
255 heads, 63 sectors/track, 522 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes


Disk /dev/dm-1: 4294 MB, 4294967296 bytes
255 heads, 63 sectors/track, 522 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes


Disk /dev/sda: 107.3 GB, 107374182400 bytes
255 heads, 63 sectors/track, 13054 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/sda1 * 1 13 104391 83 Linux
/dev/sda2 14 13054 104751832+ 8e Linux LVM
Wiping cache of LVM-capable devices
Wiping internal VG cache
Walking through all physical volumes
/dev/cdrom: open failed: Read-only file system
Attempt to close device '/dev/cdrom' which is not open.
PV /dev/sda2 VG VolGroup00 lvm2 [99.88 GB / 0 free]
Total: 1 [99.88 GB] / in use: 1 [99.88 GB] / in no VG: 0 [0 ]
Wiping cache of LVM-capable devices
Wiping internal VG cache
Finding all volume groups
/dev/cdrom: open failed: Read-only file system
Attempt to close device '/dev/cdrom' which is not open.
Finding volume group "VolGroup00"
Reading all physical volumes. This may take a while...
Found volume group "VolGroup00" using metadata type lvm2

OK. Looks good. Now (still booting from ISO or CD) if running 'vgchange -tvay' output looks OK, can you make the VG available manually with 'vgchange -ay'? If that looks OK as well, what does 'lvdisplay' return (e.g. available or not or just run 'lvscan')? If the LVs are available, are you able to fsck the device(s)?

outcome of vgchange -tvay

[root@localhost ~]# vgchange -tvay
Test mode: Metadata will NOT be updated.
Finding all volume groups
Finding volume group "VolGroup00"
2 logical volume(s) in volume group "VolGroup00" already active
2 existing logical volume(s) in volume group "VolGroup00" monitored
Found volume group "VolGroup00"
Found volume group "VolGroup00"
Activated logical volumes in volume group "VolGroup00"
2 logical volume(s) in volume group "VolGroup00" now active
Test mode: Wiping internal cache
Wiping internal VG cache


and then the outcome of vgchange -ay

[root@localhost ~]# vgchange -ay
2 logical volume(s) in volume group "VolGroup00" now active

and the outcome of lvdisplay

[root@localhost ~]# lvdisplay
--- Logical volume ---
LV Name /dev/VolGroup00/LogVol00
VG Name VolGroup00
LV UUID 3VLCHY-za4x-fDiw-17xa-PV7u-KCX8-FmBlqK
LV Write Access read/write
LV Status available
# open 0
LV Size 97.94 GB
Current LE 3134
Segments 1
Allocation inherit
Read ahead sectors auto
- currently set to 256
Block device 253:2

--- Logical volume ---
LV Name /dev/VolGroup00/LogVol01
VG Name VolGroup00
LV UUID rQ0KAR-xUJh-34oa-MSsJ-30dM-v8HL-5jrCUV
LV Write Access read/write
LV Status available
# open 1
LV Size 1.94 GB
Current LE 62
Segments 1
Allocation inherit
Read ahead sectors auto
- currently set to 256
Block device 253:3


seems everything is ok according to that logical volume scheme, with which fsck command i can try to fix my vmdisk?
fsck /dev/sda2 is the right command or not?

In your previous post fdisk output showed that you have one disk available: (/dev/sda) containing a Linux partition (/dev/sda1) and one occupied by LVM (/dev/sda2). Think of LVM as a set of "containers" if you will. You "open" each to access another. Your pvscan command showed the sda2 partition is used by one Physical Volume (PV), see this as the most outward container. One container inward, your vgscan output showed the PV contains one Volume Group (VG) called "VolGroup00" (standard naming scheme). Finally your vgchange/lvdisplay output shows the VG contains 2 Logical Volumes (LV). These could be seen as the near-innermost containers (because within each LV actually is a formatted filesystem). According to your lvdisplay output the active (by running vgchange) volume device names are /dev/VolGroup00/LogVol00 and /dev/VolGroup00/LogVol01. Run fsck on those. It could be beneficial to do a verbose "dry run" ("-v -n") fsck to show what needs correcting. That way you create the opportunity for yourself to be able to ask questions if you see something fishy before actually correcting things.

Hi again,

i made fsck -v -y /dev/VolGroup00/LogVol00 but the outcome was clean and it can't find any failure then i tried to open the server but now the old failure messages are not avalaible and there is a different failure message shows and it freezes;

Reading all physical volumes. This may take a while...
Found volume group "VolGroup00" using metadata type lvm2
2 logical volume(s) in volume group "VolGroup00" now active
exec of init (/sbin/init) failed!!!: Not a directory
Kernel panic - not syncing: Attempted to kill init!

Boot with the installer / Live CD again, go through the motions to mount both LV's. See if you can locate directory var/lib/rpm in one of them. If you have, replace DBPATH with the location where you found it and LVMMOUNTPOINT with the mount point of the LV, then run 'rpm --nodeps --nodigest --noscripts --nosignature --dbpath [DBPATH] --root [LVMMOUNTPOINT] -Vva 2>&1|tee -a /tmp/rpm_verify.log' on each. /tmp/rpm_verify.log now should contain a log to read. See 'man rpm' at "VERIFY OPTIONS" for the meaning of the results.

Yes i can reach and open the /var/lib/rpm with Live CD but could you please clarify the rest a little bit more bcs i don't want to do a mistake..

The above post does not contain commands that, in terms of regular operation, will damage your computer. It only helps you to assess damage.

Make certain that you're not reading /var/lib/rpm because that's the root of your CDROM. Your LV devices are /dev/VolGroup00/LogVol00 and /dev/VolGroup00/LogVol01. If mounted automagically grep /proc/mounts for them else mount them manually wherever you like: 'mkdir /tmp/LogVol0{0,1}; for n in 0 1; do mount "/dev/VolGroup00/LogVol0$n" "/tmp/LogVol0$n"; done'. If you do that your DBPATH will be either /tmp/LogVol00/var/lib/rpm or /tmp/LogVol01/var/lib/rpm. So if the RPM database resides in /tmp/LogVol01/var/lib/rpm then the commandline should look like 'rpm --nodeps --nodigest --noscripts --nosignature --dbpath /tmp/LogVol01/var/lib/rpm --root /tmp/LogVol00/ -Vva 2>&1|tee -a /tmp/rpm_verify.log; rpm --nodeps --nodigest --noscripts --nosignature --dbpath /tmp/LogVol01/ --root /tmp/LogVol01/var/lib/rpm -Vva 2>&1|tee -a /tmp/rpm_verify.log ' because you should run it on both mount points.