On Mon, 2005-05-16 at 23:32 -0500, Bruce Allen wrote:
> Hmm, this is certainly strange. Something you might try doing to track
> down where the problem sector lies is to do:
> dd if=/dev/hda of=/dev/null bs=512
> and then see (from stderr output or syslog) where the dd command/hda fails
> to read a sector.
Yes, this is what I finally did, and found several bad sectors that I
could remap. This is a tedious process ;-)
I re-ran the long test one more time, just to be sure and it produced a
more useful LBA_of_first_error this time (at least something that
indicated a real bad sector). So I guess this was a temporary glitch.
> Does the ATA error log show the LBA of any failed reads?
The ATA error log contained LBAs of bad sectors I manually read (and
remapped) before running the long test, but not those found by the test,
which is quite counter-productive IMHO.
Anyway, thank you for your help (and such a good sotfware).
Days of Wonder