Bad Blocks... Should I RMA?
Ronald F. Guilmette
rfg at tristatelogic.com
Thu Nov 19 01:53:14 UTC 2009
In message <44my2n45zd.fsf at be-well.ilk.org>,
Lowell Gilbert <freebsd-questions-local at be-well.ilk.org> wrote:
>"Ronald F. Guilmette" <rfg at tristatelogic.com> writes:
>
>> Nov 15 15:24:17 coredump kernel: ad4: FAILURE - READ_DMA status=51<READY,DSC
>,ERROR> error=40<UNCORRECTABLE> LBA=256230591
>
>This is *not* necessarily a big deal, despite what your other response
>told you. Errors on reads do not mean that your drive's bad-sector
>table is full; only errors on write indicate that. If you can try
>manufacturer's drive diagnostics, do that. If you can't, then it's
>harder to fix things up, but not impossible; write back if you
>really can't use a low-level diag.
Well, OK, so I ran the "long test" in Seagate Seatools.
Result: It noticed exactly and only _one_ bad sector, which I of course
told it to repair/remap.
The kicker: The specific bad secror was somewhat near to the one that
was indicated in the FreeBSD /var/log/messages errors, _however_ it was
most definitely a different LBA sector #.
So at that point I was kinda worried, cuz now I was looking at two bad
sectors.
Still, I didn't really want to have to RMA a drive that's less than one
year old, so I used dd and copied /dev/zero to the whole drive. Then
I used dd again to copy the wole drive to /dev/null. No errors at all
in either case.
Then I ran the Seatools "long" diagnostic again. Result: Zero errors.
So I decided to keep the drive. Yes, there's some bad sectors, but I
think they are all locked out now.
I'm hoping this isn't a mistake. Only time will tell.
Regards,
rfg
More information about the freebsd-questions
mailing list