|
Disk Sector errors on /
All,
I'm getting some rather nasty disk errors on my FC1 box, that rebooting/e2fsck-ing doesn't seem to help. The bonus is that it's on /...
The messages in /var/log/messages are similar to:
Dec 23 17:48:41 mymachine kernel: end_request: I/O error, dev 03:41 (hdb), sector 163352
Dec 23 17:48:41 mymachine kernel: hdb: dma_intr: status=0x51 { DriveReady SeekComplete Error }
Dec 23 17:48:41 mymachine kernel: hdb: dma_intr: error=0x40 { UncorrectableError }, LBAsect=163423, high=0, low=163
423, sector=163360
There are many messages like this, and after a reboot/e2fsck, there are many directories/files in /lost+found. running badblocks lists man, many bad blocks. I even had to tail /var/log messages to see the errors, because vi'ing messages only shows entries up to Dec 20 (it's Dec 23 now). So things are hosed, and complicated by the fact that it's / that's screwed up. And the Sector errors keep appearing, even after the e2fsck.
So, I have a few questions before I proceed:
1) I googled a bit, and saw hints that errors like this could suggest viruses. Anyone have any comments about that?
2) If it's not a virus, should I assume that the disk is hosed? i.e. no hope to eventually reformat/reuse it? It's a Western Digital 200 GB drive, and not very old as I recall, but I guess things happen.
3) Here's the really important question, the one that I'd REALLY appreciate advice on: How should I go about recovering? I have backups, so I'm not worried about data loss. But this is currently my only PC, so I'd like to minimize downtime. Is there a recommended way to recover? I was thinking about somehow plugging in a fresh drive to the existing system, reinstalling FC1 on it, and then swapping the new, good drive for the bad one, but I'm not sure if that's possible or feasible. Any ideas would be appreciated.
Thanks,
Steve
|
