Onsokumaru
Sat Aug 30 22:39:59 PDT 2008
"Alec S." <@> wrote in message news:uwxYmkxCJHA.1628@TK2MSFTNGP03.phx.gbl...
> Twayne wrote (in news:uc5natvCJHA.524@TK2MSFTNGP06.phx.gbl):
>
>> IMO defragging has no bearing on "wear and tear" on a drive; it's
>> irrelevant. If the MTBF on a disk is 5 years, you will not get ten
>> years out of it by doing exactly half of the accesses that would occur
>> in the five years. It just doesn't work like that.
>
> Sure it does. The more the head has to move back and forth, the quicker
> the
> armateur will fail and the greater the chance of the head crashing.
> Granted,
> drives today are built much better than in the past (although even that's
> debateable), but that does not change the fact that the more that phyiscal
> parts
> move, the faster they will fail.
>
> The M in MTBF stands for mean, in other words, an MTBF of five years means
> that
> /on average/, it will be five years between failures, but it depends on
> usage,
> environment, etc. If you use the crap out of the drive, it *is* usually
> going to
> die sooner than one that's idling most of the time. Of course it depends
> on the
> activity and parts involved. For example, defragging alot would result in
> the
> heads/arms dying quicker, but not the spindle, but powering up/down a lot
> *would* cause the spindle to die because of the frequent spin-up/down.
>
> Think about your body. Doing a lot of reading would wear your eyes, doing
> a lot
> of typing/writing would wear your arms and fingers, doing a lot of
> weightlifting
> would wear your joints and knees; and this is with organic systems that
> can
> repair themselves!
>
>
A very poor analogy - doing nothing will cause your body to atrophy, and
performance will be extremely low - far lower than remaining active.
You don't, "wear out", your eyes by reading, you can damage them by abusing
them, but not wear them out.
Do you, "not think", in order to save your brain?
MTBF isn't measured by running a drive till it dies.
http://www.storagereview.com/guide2000/ref/hdd/perf/qual/specMTBF.html
Of course using a device causes wear, and it will wear out eventually be
being used, but if the HDD is active the amount of extra time spent
defragging is probably so small it's not worth thinking about.
Of course if you set your machine to defrag every minute, then the amount of
time would be significant.
You would have to know exactly when the drive would fail to measure the
impact, but I doubt many people wait till the drive fails, then add up the
time they spent defragging and curse those lost hours.
It's more like saying you won't drive your car one extra mile a week because
over a year you will, "lose", fifty two miles, and you could have travelled
that much further before the car broke down.