You are not logged in.
Hi Archers,
Apologize if this is posted it the wrong place.
Today when I woke up and turned on my rig, a system enforced fsck was required. It asked me to run it manually so I did. It ran for almost 11.000 seconds complaining about all sorts of things like duplicates, illigal inodes what have you. Now when it is done to run fsck and I try to reboot I get stuck at a boot prompt ie. it just says boot: - waiting for input. There is an error :
Failed to load libutil.c32
Failed to load COM32 file menu.32
boot: [waiting for input]
I have no idea what is going on, I've never been in a situation like this before.
I tried to see what LS would do and it looks like the files it complains about are there, what to do?
I can just add to this that I have not done any recent updates.
UPDATE 11/07
I made a bootable usb with arch an fired it up.
when trying to mount the partition:
EXT4-fs (sdb1): ext4_check_descriptors: Checksum for group 59 failed (47268!=44979)
EXT4-fs (sdb1): group descriptors corrupted!
mount: wrong fs type, bad option, bad superblocks on /dev/sdb1, missing codepage og helper program or other error.
syslog says:
sd 9:0:0:0:0: [sdb] tag#19 Sense Key : 0xb [current] [descriptor]
sd 9:0:0:0:0: [sdb] tag#19 ASC=0x0 ASCQ=0x0
sd 9:0:0:0:0: [sdb] tag#19 CDB: opcode=0x2a 00 00 00 00 20 00 00 08 00
blk_update_request: 12 callbacks suppresed
blk_update_request: I/O error, dev sdb, sector 32
buffer_io_error: 328 callbacks suppressed
Buffer I/O error on dev sdb1, logical block 0, lost async page write
ata10: EH complete
EXT4-fs (sdb1): ext4_check_descriptors: Checksum for group 59 failed (47268!=44979)
EXT4-fs (sdb1): group descriptors corrupted!
Last edited by odp (2015-07-14 09:39:14)
Offline
Regarding the bad supeblocks, I tried to to change to backup superblocks but they also seems to be corrupted.
I am changing the title on the question to something more suitable, hope moderator will move to the post to a more appropriate place if needed
Offline
Well, I've never seen a fsck that took longer then a few seconds do any good so I think you should reinstall from backup. If you want to, you can attempt recovery instead; but I suggest no longer mounting the filesystem writable.
Also; the i/o errors suggest hardware failure to me so I'd have a good look at your hardware and consider replacing the hard-drive. Fwiw, the errors come from syslinux, and since the filesystem is corrupt that doesn't surprise me
Offline
+1 for consistent pattern with hardware.
1) Backup if you haven't already
2) `smartctl --test=long /dev/sdxy
3) e2fsck -cckv /dev/sdxy
Offline
I had a hard drive a few years ago that kept losing it's super blocks. It happened three times in a period of two months and then I tossed the drive. I haven't had that problem since.
Offline
Thanks guys!
I found an old drive and did a dd clone of the drive so I didn't have to work on the failed disk.
Then went out and got me a 850pro sdd disk to start building my system again.
e2fsck -cckv /dev/sdb1 did a run in about 5 hours and after that I could mount my disk again, I was able to recover most of my stuff which was real nice.
One file that I needed to restore mysql databases was missing, so now I am doing a few extra runs to see if I can recover it!
Thanks you are the best!
Offline