[Cialug] Google Hard Drive Study

[Nathan Stien]

On 3/8/07, Tim Wilson <tim_linux at wilson-home.com> wrote:
> Ok, but correct me if I'm wrong, if they have such a good working
> relationship, wouldn't hard drive manufacturers make sure Google get
> the primo quality drives?  Just to make sure their relationship with
> Google wasn't ruined?  When you've got a show dog, you don't give it
> the same chopped "whatever" you give every other dog.

But how do you know which ones are the primo quality ones?  Beyond the
basic factory checks and S.M.A.R.T. stuff, you don't really know which
ones are going to turn out to be lemons in six months.  You can do
some initial burn-in testing to cut down on infant mortality, but that
still doesn't distinguish a 6-month drive from a 6-year drive.  Also,
even if you did that, you'd still have to account for the fact that
Google will do their own burn-in before putting them into service.

It is possible that if they do have some clever way of sifting out the
extra good drives, they are already doing so and the data from the
study are therefore skewed, meaning our drives are even crappier than
this document would suggest. ;-)

-- 
Nathan P. Stien
Consulting Engineer / Software Developer
Embedded Systems Electronics and Software
http://linkedin.com/in/nathanstien
Mobile: 309.241.2581