EXT4-fs group descriptors corrupted
Hello All,
My hard drive recently failed on my raid and got a new drive and got the server booting again, but now my NFS 2TB drive won't mount.Can you help please. Thank you Code:
lvs Code:
mount -t ext4 /dev/vgdata/lvdata /data/ Code:
mke2fs -n /dev/sdb1 Code:
fsck /dev/vgdata/lvdata Code:
--- Volume group --- Code:
lvm lvs -a -o+devices |
Did you google the error message "device doesn't appear to be in the dev hash table"?
Are you sure that lvdata has an ext filesystem? Not xfs or something else? Try running blkid; it tries to figure out what is on that volume. Also, this: Code:
mke2fs -n /dev/sdb1 Finally, to clarify, is lvdata located on your RAID? Are there other volumes on the RAID that have problems? What kind of RAID, hardware or software? Any messages referring to the RAID and lvdata in the message buffer (dmesg or /var/log/[syslog|messages])? |
Thank you for your assistance.Yes, I did an extensive search in google.
Yes i am sure it is EXT4 Code:
dev/mapper/vgdata-lvdata: UUID="f4803c5a-8149-438b-a16e-fbd2a73c9ffb" TYPE="ext4 I have hardware RAID, there is no reference to the device in dmesg Code:
XT4-fs (dm-2): ext4_check_descriptors: Checksum for group 3992 failed (18287!=0) m Code:
ke2fs 1.41.12 (17-May-2010) |
Quote:
Code:
/dev/sdb1 is apparently in use by the system; will not make a filesystem here! Code:
fsck /dev/vgdata/lvdata You can also try fsck -b <one of the backup superblocks displayed by mkfs>, but it seems that it does it by itself already, with little success. |
Yes it is relevant because the system is on a raid array and on that raid I use vcenter to make a host then from that create a vm, and assign hard drives.There are two hard drives in the system, /dev/sda the root and boot partition and /dev/sdb the nfs storage.
You are correct, fsck, never attempt to repair anything it just throws an error.I did try in single user mode, but will try again after the drive has finished being imaged. |
1 Attachment(s)
I tried running FSCK in single user mode now, and I am getting this error in the screen shot.
|
I just ran fsck again, and again it deletes my data.why is that? there is nothing in lost+found except 2 files that are 400mb in total.The size of data on there was 2.TB.There is nothing in the logs either
|
I guess nothing?
|
Quote:
It's also hard to say what's going on. Why lvdata is "busy" although you are in single user mode is a mystery to me. Regarding the data: fsck repairs filesystem structures, not data. It identifies inconsistencies like wrong summary counters (benign), files that have no directory entries (will be moved to lost+found) and data blocks that belong to no files but are not marked free either (will be marked free). File content is not considered. So you may say that fsck has removed your data, but what it has really done is marking unaccounted-for data blocks as free, as it is unable to guess what files they might belong to. So after fsck has done its work, you may end up with an empty, but consistent filesystem. Your data is probably still there. With the help of a filesystem debugger like debugfs you might be able to salvage something, but it would be very tedious. My feeling is that something is wrong with the way you defined the physical and/or logical volume(s) on top of your RAID, but I would have to play around with the disks to get clues. |
All times are GMT -5. The time now is 09:21 AM. |