On Mar 15, 4:35 pm, "Folkert Rienstra" <see_reply...@myweb.nl> wrote:
> <pau...@efn.org> wrote in messagenews:1173925752.221642.270340@e1g2000hsg.go oglegroups.com
> > On Mar 14, 6:40 am, "Folkert Rienstra" <see_reply...@myweb.nl> wrote:
> > > <pau...@efn.org> wrote in messagenews:1173866711.145155.278150@y80g2000hsf.g ooglegroups.com
> > > > On Mar 13, 3:14 pm, Jax <inva...@no-mail.com> wrote:
> > > > > Home user with XP Pro.
>
> > > > > I bought some 160 GB hard drives a couple of years ago and they got
> > > > > filled up with data.
>
> > > > > I have now migrated all the data off these 160 GB hard drives and will
> > > > > now use the drives to hold backups.
>
> > > > > QUESTION ---> As the HDDs are now empty is it worth writing zeros,
> > > > > before using them again, in order to force the HDD to map out any
> > > > > defective sectors?
>
> > > > > QUESTION --> Or will mapping out of any defective sectors happen
> > > > > automatically when any bad sectors are next written to, which means it
> > > > > is not worth writing the zeros?
>
> > > > It's worth zeroing the data on the drives under two circumstances.
>
> > > > 1) You're selling them, and don't want any one to steal your bank
> > > > account info, etc.
>
> > > > 2) You don't want anyone to see your porn collection :)
>
> > > > If it has bad sectors, replace the unit. According to a very recent
> > > > google study of over 100,000 consumer grade hard drives,
> > > > those with read errors were 39 times more likely to fail within
> > > > 60 days than those without.
>
> > > Which says absolutely nothing if those without don't fail, now is it.
>
> > Of COURSE drives without bad sectors do fail.
>
> But not all. Without a percentage, 39 times or 16 times is a useless number.

The percentages are on figure 6.



>
> > Otherwise the statement would be different.
>
> > > And the exact phrase was:
> > > "After the first scan error, drives are 39 times more like-
> > > ly to fail within 60 days than drives without scan errors."
>
> > > Unfortunately there is no such thing as a 'scan error'.
> > I'd presumed it was a synonym for a bad spot on the drive in question.
>
> Yup, detected by a very particular action of the drive. Problem is,
> there is no such attribute with that name. So how will they know.
> They particularly isolated them from the pending ('probational')
> counts, the online and the offline reallocated counts, so it's not those.
>
>

It's also not Seek error, or CRC error (listed separately).

It's most likely that they are referring to attribute 1. (Read Error
Rate)

http://en.wikipedia.org/wiki/Self-Mo...ing_Technology



> > > Your 'read' errors appear under probational counts.
> > > "The critical threshold for probational counts is also one:
> > > after the first event, drives are 16 times more likely to fail
> > > within 60 days than drives with zero probational counts."
>

They're different.

Probational counts is attribute 195.


> > > There were other inconsistencies in the report as well, like lower
> > > risk numbers for the total lifetime (longer than 60 days).
>
> > > They also didn't say what they considered a failure and whether the
> > > 'failed' drives actually failed in a different system once replaced.
> > > Neither did they check whether it was the system killing the drives.
>
> > Actually, yes they did specify what they counted as a failure. I'm
> > paraphrasing because I don't have the report in front of me. "A drive is
> > considered to have failed if it was replaced as part of a repair operation".
>
> But no explanation of what a 'repair operation' is and what prompts it.
> A simple single bad block in the wrong place may prompt a 'repair operation'
> where the drive is simply replaced as part of a quick fix by lack of other repair
> options. That doesn't necessarily mean that the drive itself is beyond repair.
>
A repair operation is a pretty broad subject.

Not many people bother to fix failed drives, except for the purposes
of data recovery. Though it's possible Google does. I would however
consider a drive that needed to be repaired to have failed.

By the underlying tone of the report, replaced drives were often
retested.

"From an end-user's perspective, a defective drive is
one that misbehaves in a serious or consistent enough
manner in the user's specific deployment scenario that
it is no longer suitable for service. Since failures are
sometimes the result of a combination of components
(i.e., a particular drive with a particular controller or cable,
etc), it is no surprise that a good number of drives
that fail for a given user could be still considered operational
in a different test harness. *** We have observed
that phenomenon ourselves, including situations where
a drive tester consistently "green lights" a unit that invariably
fails in the field.***"


You can also tell that today I DO have a copy of the report in front
of me. :)


>
> > > > It's worth checking the drive for bad sectors.