IBM Deathstar Hell

[Monty J. Harder]

"Paul Taylor" <paul at kcnetcare.com> wrote:

> Bad sectors happen to magnetic drives and is no indication the drive is
> faulty unless they fail often or a lot. IDE, SCSI, or SATA is not

  But they also transparently substitute good sectors from a reserve pool,
so that the computer should never see the bad sectors until the pool has
been exhausted.  At that point, the drive is usually not long for this
world, and needs to be replaced ASAP, because the 'grown' (excluding the
initial 'infant mortality' ) defect rate tends to rise with age.  The LLF
that was originally described forces the drive to create a whole new defect
list, then set aside some more good sectors in reserve.  The fact that soon
after, those good sectors are gone, is proof that the drive continues to
grow defects daily.

  I bet if you take a close look at the drive after each LLF, it's reporting
a slightly smaller capacity than the time before.