Kevin Shackleton <[email protected]> writes:
> On Tue, 2009-01-13 at 13:59 +1100, Daniel Pittman wrote:
>
>> Actually, temperature has very little relationship with disk life, at
>> least when Google studied their consumer grade disk failure metrics.
>>
>> The details: http://research.google.com/archive/disk_failures.pdf
>
> Thanks for that link Daniel. An interesting paper. However, I don't
> know that I'd compare a server farm environment to home PCs.
That is fair: I know, for me, that my hard disks run at comparable
temperatures to the temperatures they mention, but I guess I have seen
enough poorly designed cases with fancy looks that I could be on the
uncommon end of things.
> My gut feeling from the tens of hard drive failures I've worked on is
> that close-stacking drives is a bad thing. Maybe because
> through-ventilation in big home boxes is not as thorough as in rack
> box servers.
*nod* Certainly, airflow can make a difference. Checking now, in the
heat, the newer box I own — which is in a better designed case — is
around 7°C cooler on average across the hard disks.
I hadn't actually noticed that there was that much difference in extreme
temperatures. Ah, well.
> Another bad thing is what happened to me a couple of days ago when the
> power supply variable speed fan went to low speed (not stopped) and I
> smelled the PC getting hot. I removed a little control circuit board
> from inside the unit and the fan now runs at full speed permanently.
> Fortunately nothing died (so far).
*nod* Good luck.
Daniel
--
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
