28 Nov
2010
28 Nov
'10
6:59 a.m.
On Sat, Nov 27, 2010 at 7:24 PM, Eric Smith <eric at brouhaha.com> wrote:
Chuck wrote:
I can echo this, as I have ran servers for years straight, without a hard
drive failing, but I have pulled drives off the shelf and have them fail on
power up. one must also account for manufacturing defects that can manifest
themselves for any amount of time before causing a failure. environment also
plays a role, as simple events such as seasons changing between extreme hot
and extreme cold changes the structures of metal bearings with very tight
clearances, causing friction easily, and humidity changes along with
condensation also have an effect on drives. I am sure, even in a slight
vacuum at 72 degrees constant and being spun up every month or so a drive
would still fail.
So time spent spinning is quite germane in a disk
drive.
Obviously.
However, there are failure mechanisms that are not related to spinning, so
having an old drive with few operating hours is NOT a guarantee that it will
work reliably now.
Anecdotal evidence suggests that leaving a drive non-operational for a long
time is bad, and that spinning it for a few hours a month may actually
extend its life.
Eric