Google HDD Research

[Plotloss]

http://news.bbc.co.uk/1/hi/technology

Quite interesting, if you're a bit sad, like me.

Keeping them warm and cosy and used constantly means a longer life for HDDs

[GregE240]

Surprising results....

Less use means more failures?

[sneakyneil]

Seems they stopped short of breaking it down by model/manufacturer though...

And 3 year old drives are the most reliable?

[chilled]

if a drive lasts 3 years, it's proven itself to not have most of the early failure faults.

But yes, one of the biggest causes of disk failure is a disk laying unused for too long, then failing to spin up. A lot of hard drive manufacturers have added firmware to keep drives spun up.

[Zod]

interesting.

I had two hard drives fail in a few months in one PC and found that they were getting very hot. I cut a hole and installed a fan and two years later have not had enother failure.

[Lord Grover]

I've not read the whole article, so may be repeating what's there, but in my experience it's long-lived, constant use drives that are prone to failure. I've always assumed it is because they are spinning more or less constantly that the bearings/spindles wear - which is fine while they are still spinning - but as soon as they stop due to a real shutdown (not restart) they 'fall off' the spindles never to spin again. It's happened to me sveral times over the years.

[mph999]

Lord Grover said:

I've not read the whole article, so may be repeating what's there, but in my experience it's long-lived, constant use drives that are prone to failure. I've always assumed it is because they are spinning more or less constantly that the bearings/spindles wear - which is fine while they are still spinning - but as soon as they stop due to a real shutdown (not restart) they 'fall off' the spindles never to spin again. It's happened to me sveral times over the years.

I'll agree with that ...

We have 1000's of drives at work (not joking), on the very rare occassion we have a real shutdown, it's a real worry as we know things will not come back to life ...

Best bet, keep em spinning ...

Martin

[agent006]

The only time we've had a harddrive go in a desktop at work is after the christmas shutdown every year. Without fail one or two machines won't come back.

[roadsweeper]

GregE240 said:

Surprising results....

Less use means more failures?

Even Google described them as surprising and I would agree with that: Hard Disk Test 'Surprises' Google. Have a read of the article and it will describe reasonably concisely in what area the results were unexpected.

This is probably the largest and most detailed study into this area so the results are probably more representative than anything seen to-date.

[Carrera2]

Without reading the article linked (I read a bbc one) - does it also mention the survival of the fittest theory? I.e. that the badly made ones fail early and the top quality ones are still going at the end of the test.

A difficult product to produce an accurate and unbiased control model of I suspect.


Eta - I've jsut noticed that this is the bbc one. :doh:

Edited by Carrera2 on Wednesday 21st February 11:04

[rich1231]

any link between the amount of pron on the disk and longevity? If so Shady's disks are going to last for eternity.

[sneakyneil]

rich1231 said:

any link between the amount of pron on the disk and longevity? If so Shady's disks are going to last for eternity.

My own experiments have found no such link