On Tue, Jan 27, 2004 at 11:39:36PM +0100, Jan wrote: > Hi, > > I'm having troubles with an old (but new as in 'never used before by me') hd. > > Some facts: > -The HD contains a very small redhat system: I was able to install it using a > cd-rom-player and I start the gentoo system install from chrooting out of redhat. > -It's a WD Caviar 21600 (1.6 GB) > -The system is a ibm pc 350 (PI 100Mhz, 32MB ram) > > The problem: > - The kernel keeps printing out I/O errors, I noticed because 'emerge sync' never > finishes ok. > - I've started doing 'badblocks' runs. But seen that it complains each time about > other bad sectors. > - After some thinking, I've tried to write the partition table again (with fdisk) > because I remembered a problem with fdisk writing the table:
Here is what I'd do: Boot from a 'clean' disk (i.e. gentoo install CD, Knoppix, etc). Run 'destructive' badblocks test because non-destructive ones (i.e. the ones that don't try to write aren't as good at catching the problems). This WILL destroy ALL the data! If it seems happy (no bad blocks)- use it, if not - it's probably time to replace it. Judging by it's size, it's old. It's been my experience that when kernel starts printing IO errors in a log file, hard drive is dead. Usually what happens next is 'clicking' sound and hard drive dies completely. I also don't recommend using a hard drive with 'just a few' bad blocks. AFAIK modern hard drives have a built-in 'self healing' mechanism and you only start seeing errors when it can no longer 'heal' itself (i.e. reallocate bad blocks). So even if it's not dead, it's probably very close to it at that point. -- - Andrey ~ In theory, practice and theory are the same, but in practice they are different (Larry McVoy) ~ -- [EMAIL PROTECTED] mailing list
