On 03/14/2018 06:30 PM, Computer Nerd Kev wrote:
Percival P. Cassidy wrote:
On 03/13/2018 06:09 PM, Computer Nerd Kev wrote:

Just wondering. My friend does. I wonder if I should do that too.

Thank you in advance.

No. The most likely source of failure is one of the chips. There is no
way to do a proper stress test on the electronics of a device.

Well there is, but to do it well might require removing the chips from
the HDD's circuit board and building a custom circuit to stress test
them. You generally assume that the manufacturer of the chips has gone
through enough testing to be reasonably confident that they conform to
their specifications. I'd imagine the same for the mechanicals of
mechanical HDDs.

I'd say that if you need to stress test your HDDs, then you don't
trust the manufacturer, in which case that prompts the question of
why you bought from them in the first place.

Many years ago I read that modern solid-state electronics either fail
within the first 100 years or run for a very long time. And a hard disk
is not entirely solid state

Ooops! I meant "100 *hours*" of course.

Perce

They can last for 100 years, or if they are operating at the limits of
their temperature rating (as is sometimes the case in poorly designed
equipment, or if it is operated in a harsh enviroment (for a HDD - not
enough airflow)) they might last a much shorter time. There will also
be some that, when made, just don't turn out quite right and will fail
in a _very_ short time. However as I stated, the manufacturer of the
electronic components should have sufficient tests in place to make
sure that the dodgy chips are weeded out before they get anywhere
near the HDD manufacturer.

I certainly expect that the HDD manufacturer would then run a burn-in
test on their finished HDDs to make sure that they are in good order
before shipping them off. If so, doing it again would be a waste of
time.

As for "trusting the manufacturer," if you read user reviews on line,
you will find that, no matter which manufacturer, there are always
people who will never again buy drives by that manufacturer. IOW, every
manufacturer ships the occasional dud.

I therefore run extensive tests on all my new drives.

After the manufacturer has stress tested the drive, I would expect that
further early failures would be from some component that simply breaks
after extended use rather than not being built to the standard required
to survive a high load situation. The only way to find that out is to
just run the drive and see how long it lasts - little point to extra
stress tests.