clipka wrote:

> Stefan Viljoen schrieb:
> 
>> I've seen this phenomenon on older drives. Had a HDD in a system here
>> that was left on for years, probably been turned off 20 or 30 times in
>> its useful life. The fortieth or forty-fifth time the server was turned
>> off, the HDD died - after running reliably (while not being turned off,
>> ever) for 5 or 6 years. Ok, old hardware and an old drive, but just goes
>> to show. Keep 'em spinning is better than spin-up / spin-down on a daily
>> basis.
> 
> To me that rather makes a case that running a HDD 24-7 nonstop (probably
> at high workload) may instead just increase the risk of
> death-by-power-cycle.
> 
> Otherwise you'd wonder how office computers, being powered on and off 5
> or 6 days per week, could possibly last even two months. Yet they're
> usually good for about 5 to 10 years (though they'll be long outdated by
> then).

Hmm, yes, but I've often wondered if there -really- is a difference between 
"desktop hardware" and "server hardware" - and that maybe they are designed
for different MTBF? I. e. a "server grade" HDD will last longer if not
regularly turned on and off, while a "desktop grade" HDD will die quicker
if never turned off?
-- 
Stefan Viljoen

Post a reply to this message