Can't even get into the rescue disk
P4 running Ubuntu 8.10 with 4 physical disks, no RAID. One of the non-system disks, /dev/sda with ext2 (I think, it was from an old system that I mounted), is giving me read errors while extracting some rars. Unmounted disk, run fsck -c /dev/sda which takes all night. Decide to kill the fsck process this morning as when I first started it, it had said "read-only test" so I (wrongly?) assumed that it wasn't actually writing anything, probably a bad idea, but seemed brilliant in my uncaffeinated state when I got up. :/
Anyways, rebooted system, chokes on "Alert! /dev/by-uuid/some-guid-to-bad-drive" error, so was hoping to be able to get into a shell where I could edit fstab and comment out the bad drive so I could boot up system and continue trying to fix/recover the disk. However, I can't seem to get anywhere off of any live cd.
With Rescue Remix, it gets to the options menu (memtest, enter to boot, etc.), hit enter, SRST failed (errno -16), continues, then get buffer I/O error, end request, etc. for sr0 (which I'm assuming was sda?), which I believe is the bad drive.
bad controller? Kind of clueless as how to proceed. Hoping to salvage as much data as I can, as I have another equivalent-sized drive mounted/available to copy the data to.
Probably the shortest path
Probably the shortest path to take would be to remove the bad drive from the system and boot. Then, image the drive using a usb-to-IDE or USB-to-sata controller.
Take a look in /var/log/messages as you plug the USB controller in the computer to see if the drive can actually power up properly.
I actually went in through
I actually went in through the Grub menu (chose the "hd" boot option to boot from the first disk) and the system booted after a few CTRL-Ds. I've now since shutdown, pulled the drive and strangely, it seems to be reading fine from a mount via USB enclosure.
However, during boot, I'm still getting a bunch of:
Buffer I/O error on device sr0, logical block 40124
[14.230695] end_request: I/O error, dev sr0, sector 320992
About 9-10 instances in dmesg, all same block, dev, and sector. Not even sure what sr0 is; I pulled the other data drive so now there's only the two SATA drives hooked up with the one "bad" drive on sdc (USB).
Stupidly, I think I commented/changed the fstab options a few months back before fsck at boot was taking HOURS and needed to get the system up to do something thinking I'd change it back, reboot, and let fsck do it's thing at a later date but of course I forgot all about that...
Now, I still think there may be something wrong with the now-USB-mounted drive on /dev/sdc: what's the proper way I should proceed to check it, repair it, etc?
Ok, smartctl
Ok, smartctl says:
Reallocated_Sector_Ct 0x0033 001 001 036 Pre-fail Always FAILING_NOW 59990
Have another identical 320GB drive along with this drive now physically mounted in the case. I'd like to backup as much data as possible: should I be using dd to accomplish this?
Boot from a Live disk and
Boot from a Live disk and use gddrescue to image the drive. It sounds like the other drive may be having problems, too. Do the same with it. Again, swap out the drive to troubleshoot the controller.