[CentOS] OT - Offline uncorrectable sectors

William L. Maltby CentOS4Bill at triad.rr.com
Fri Aug 22 16:22:26 UTC 2008


On Fri, 2008-08-22 at 08:59 -0700, nate wrote:
> Lorenzo Quatrini wrote:
> > I have few disk that have offline uncorrectables sectors;
> 
> Ideally it should be done using the manufacturer's tools,

Second that!

> and really any disk that has even one bad sector that the OS
> can see should not be relied upon, it should be considered a
> failed disk. Disks automatically keep spare sectors that the
> operating system cannot see and re-maps bad sectors to them,
> if your seeing bad sectors that means that collection of
> spares has been exhausted. I've never seen a disk manufacturer

?? Uncertain about "spares has been exhausted". I recently had one SATA
drive that kept reporting a bad sector (actually grew to three). Being
inured against panic attacks by long exposure to panic-inducing
situations, I decided to let it ride a bit (it was an empty re-used
partition upon which I would mke2fs and temporarily mount and use) and
see if the number continued to grow. To this end, I ran the smart tools
extended tests, several times over a period of a week, and saw no new
ones. This was reassuring as traditionally if failure is imminent the
number tends to grow quickly. A few appearances of bad sectors early in
the drive lifetime is not an unusual occurrence and is not reason for
trade in of the drive (after all, in this case the manufacturer just
runs the repair software on it and re-sells it). It *is* a reason for
heightened caution and alertness, depending on your situation.

After deciding the drive was not in its death-throes, I downloaded the
DOS utilities from the manufacturer web site and ran the repair
utilities. No smart tools reports of bad sectors since then (about 2
months so far).

Now, I don't know (or care) if an alternate sector was assigned, just
that the sector was flagged unusable. For my use (temporary use - no
permanent or critical data) this is fine. Last several mke2fs runs have
produced the same amount of usable blocks and i-nodes, so I don't see
evidence that no spare was available.

I do expect that a few more sectors will be found as the drive ages
until the manufacturing weak areas have all aged sufficiently to cause
failures.
> not accept a disk that had bad sectors on it (that was still
> under warranty) in as long as I can remember..

If your application is critical and you still have warranty, the only
cost is inconvenience, delay and more work to get it exchanged. You will
likely receive a "reconditioned" drive though. So for me, in my
situation, the download and use of the manufacturers repair software is
better. Only bad part is instead of using floppies now, they seem to
want a CD/DVD to boot from. A minor inconvenience considering the
alternatives.

> 
> nate
> <snip sig stuff>

HTH
-- 
Bill




More information about the CentOS mailing list