Greg Freemyer wrote:
Having an issue I thought I knew how to address, but it is not working. OpenSUSE 10.2 is booting into a maintenance mode due to a bad sector read on the hard disk. I'm not worried about the data on the drive and so far it is just one bad sector.
So I would like to cause the drive to map the sector to one of its spares, then do a fsck on the partiion, then some kind of installation verification step, or if truly necessary I can reinstall.
I'm stuck on the first step. I thought writing data to the sector would cause the drive to remap it, but that is not happening.
dd if=/dev/sda of=/dev/null skip=312518818 count=1
Is very slow to return and triggers the bad sector error messages.
dd if=/dev/zero of=/dev/sda skip=312518818 count=1
returns very fast and and reports success, but redoing the above read again still returns errors.
Is there a different way to get that sector remapped?
You should check what the smart log of that disk contains, you might get some interesting info about what is causing it. Perhaps you could start its selfdiagnostic. But maybe it would be wiser to do a backup of the affected partition first. Then, after the diagnostic, write over the whole partition and reformat it. Another idea: the remapping sectors are not infinite: if spent, remapping will fail. The smartctl tool should tell you that. It is of course a signal to immediately replace the drive. -- Cheers, Carlos E. R. (from RC1) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org