Is it possible to "wear out" a hard drive?

[Venesectrix]

I usually defragment my drive about once a week to keep my comp up to speed, but I've been hearing that frequent defragmenting "wears out" the hard drive and leads to a shorter drive life span. Since I've never seen actual tests or proof or anything, I've been kind of skeptical, but many people I find online keep giving advice like "just defragment once a month or you'll burn out your drive." So, is defragmenting harmful to the drive/does it wear the drive out?

[tom10167]

That's a bunch of crap, that would be like saying taking a shower slowly kills people because you're destroying cells.


Harddrives have a "lifespan" I think it's 100,000 hours or so(12 years running nonstop) but that's it.

[lostshoe420]

Yeah, but your example Tom, however ludicrous, still sort of makes sense. However, the miniscule amount it affects the person (or the hard drive) is insignificant in the long run.

[tom10167]

All of my analogies are ludicrous, but they get the message across in some form or another. :D :D

[shortyes]

Harddrives have a "lifespan" I think it's 100,000 hours or so(12 years running nonstop) but that's it.

Dang where can I get one of these HDD :D

but seriously, can you wear out a HDD? yes but after many many years or use, the disk or the motor will die just like all other components of a computer.

Think of a car, someone off the street sees that you are getting an oil up change. He comes ou and says that frequent oil change will wear out your car.

Now will defrag wore out a HDD? Please point me to the person that said this so I can sell them a bridge made of gold.

[Th0r]

Sorry...Thats the most stupidest thing i have heard this week... LoL... sorry but they wouldn't make tools like they have if it would shorten its life span...

[Viper848]

i have to agree with everyone else in this thread. I had the same computer for 5 years, most of the time running it at 700MHz and around the clock, oh and it only had like 250MB left on the drive. Trust me you cant wear out a hard drive, maybe after 12 years it would burn out or something. But if you were running the same PC for 12 years around the clock, I think another component would burn out before the hard drive.

And to conclude, florida will ahve a blizzard, and all the snow on the north pole will melt, before you wear out your hard drive by defragmentation.

[shortyes]

don't forget the flying pigs

[DayUSeX]

i dunno i might have to argue you guys on this one.

Had to death starts both 40gb models. The main one i defragged weekly as it had the OS the other 40gb was storage defraged montly or as needed (which was mostly montly). Main desthstar dies (doh get a wd), but stupidly keep the other death start. That thing dies i belive 6 months later. Dunno who cares down with ibm hard drives =P

[tom10167]

I think it's a given that we're talking about non-defective products. :)

[JigPu]

I can't immagine defragging (or any kind of "heavy" hard drive use) to be any more detrimental to the drive than being idle.

When defragging, the only moving parts are the motor/spindle/platters and the head/arm. There's no difference between idle and defragging platter speed, so it's not going to kill that sooner. The head/arm is moving more but it dosen't use a motor to move it (uses magnetism), so the only thing that theoretically is being weared more is the pivot point of the arm... :p


JigPu

[MadSkillzMan]

Friend had an IBM drive that died. Im sure our power has all gone out and no harm done. August 14th here in USA the famous blackout happens.....my 3 drives unaffected. His drive comes back clicking. It was out of warranty. Opens it up, and watches the arms try to move to the mid of the drive then snap back.

We looked on the IBM site. Said that IBM doesnt recommend using the drives in these space saver computers for extended periods of time. Their definition of extended was 12 hours a day.

pathetic huh?

He inherited this computer form his uncle, was one for those computers for people that hardly use one. Bout the size of a cable box.

[jlin453]

Exact question answered here:
http://www.ocforums.com/showthread.php?s=&threadid=281381



Myth #4 :


Defragmenting the hard disk will stress the needle (head actuator).

Truth :


That is actually contrary to the truth. Defragmenting the hard disk may involve a lot of seeking as the hard disk rearranges its data in a contiguous fashion. This allows the read/write heads to read large amounts of data without seeking all over the platters.

However, after defragmentation, the hard disk no longer needs to seek all over the platters for your data. This reduces the amount of head actuator movements as well as greatly increase the hard disk's read/write performance.

Therefore, while it may be technically correct to say that defragmenting your hard disk will stress the head actuators, the truth is defragmenting your hard disk will reduce the amount of seeking from then on and thus reduce the head actuators' workload.

[Khasra]

Gotta love that quote. To paraphrase:

--------------------------------
Myth #4 : Defragmenting the hard disk will stress the needle (head actuator).

Truth : It is technically correct to say that defragmenting your hard disk will stress the head actuators.
--------------------------------

Some myth eh? :confused:

Excessive defragmenting, like doing any other read/write activity excessively, will negatively impact the lifespan of a drive. The point is to find the balance where the improvements gained by an arbitrary defrag outweight the wear caused by that given defragment process.

[JigPu]

Originally posted by Khasra
Some myth eh? :confused:

Excessive defragmenting, like doing any other read/write activity excessively, will negatively impact the lifespan of a drive. The point is to find the balance where the improvements gained by an arbitrary defrag outweight the wear caused by that given defragment process.
Not really... The more you defrag, the less the head moves IN a defrag, eventually to the point where you'll run defrag and the head won't move at all since the drive is already defragged. Combined with the fact that having a constantly defragged drive reduces the amount of seeks the head has to do when accessing a file, "excessive" defragging may not be that bad.


Why defragging daily causes less seeks on the drive compared to defragging monthly:
If I defrag daily, I end up having to (usually) move around something like 200 fragments. This amounts to a few hundred seeks (at least 400). Now, the seeks accumulated from having to access the files would be rougly related to the % files fragmented. If we assume the HD is forced to seek to 10% of those fragmented files, it's another 100 seeks. Total seeks/day over a never-fragmented drive: 420 (400 defragging seeks + 20 seeks to fragments before you defragged).

If I defrag monthly, I end up having to (usually) move around THOUSANDs of fragments. It's been a week since my last defrag, and I've currently got 2K fragments built up. Assuming fragmentation occurs at a linear rate, each month I'd have 8K fragments. This is at LEAST 16K seeks, or reduced, 533 seeks/day. If things were truly linear the number would be the same, so this just shows how much a difference (33%) the real world has on numbers. If we apply the same formula to determine the number of seeks made from having to use fragmented files in the month, there's an aditional 15802 seeks in a month, or 526 seeks/day. Total seeks/day: 1059 (150% more than daily defragging :eek: )


~~~~~~~~~~~~~~~~~~~~~
But even all that is rather useless since (like I posted above), the head dosen't get weared by seeking. If you've ever ripped apart a HD before, you'll see that there is nothing to be worn by head movement. The "head actuator" consists of two very powerful magnets (great fun for playing with after you've had your fun with the drive :D), and a plastic piece attached to the arm. This plastic piece has a coil of wire on it. By running electricity through the wire, the magnetism has an effect on the magnets in the drive, and forces the plastic pice (with attached arm) to move.

The things being stressed are:
- Whatever produces the current to go through the coil
- The junction between the plastic and the arm
- The pivot point of the arm

None of which are under any appreciable amount of stress. Increased head movement WILL wear them, but it would take an order of magnatude more seeking before a noticible difference is made.


JigPu

[ajrettke]

No mechanical device can handle cyclic loading indefinetly. The question is wether or not it will truly impact the life of your hard drive. I'm not sure I really like JigPu's example, just seems way too simple to assume everything is linear. You'd need to get probabilities in there and start cranking a number out like that anyway.

The thing I'm wondering though....If you defrage a drive that builds up X framents per day, and it takes Y amount of time to defrag X fragments....The defragging once per year is gonna take just as long on a drive as defragging once a day for the entire year....
365*X fragments will take 365*Y time....whereas X fragments will take Y time...but 365 times....

[JigPu]

Linearity is a reasonable approximation, but some of my other numbers suck. Like the 10% I have was totaly arbitrary. I originally was going to use 50%, but it made the point WAY too well and I don't know if it's a reasonable guess. 10% is the smallest "reasonable" amount I could imagine, so I used that. I need better data (or time to extrapolate the basics to better rules) for any really good data. :(

JigPu

[druidelder]

Originally posted by tom10167
...Harddrives have a "lifespan" I think it's 100,000 hours or so(12 years running nonstop)...

That is not technically correct. The way hard drive manufacturers report this data is much the same way they report disk size. Misleading.

If a drive is listed with a failure rate of 100,000 hours, this does not mean that the drive shouldn't fail for 100,000 hours. It is a guage of the entire product line and not the individual drive in and of itself. The correct interpretation of this data is for every 100,000 drives, one will fail every hour. This seems like a silly distinction, but it brings the drive life down considerably. By this interpretation, if the curve maintains it's theoretical peak (which never happens), one drive out of 100,000 would make it to that life span. Around half would have died by 50,000 hours (still a long time though). This rate can be known as the "Mean Time To Failure", or "Mean Time Between Failure", etc.... Again though, it is an estimated number based on testing before release, and after release this number is altered by RMAs.

A much better rate to determine the life of drives is the annualized drive failure rate. Industry wide (meaning individual manufacturers numbers may be a little better or a little worse), the annualized drive failure rate is about 1%. Doesn't sound too bad does it? Well, it is. This means that one out of every hundred drives fails a year. Not bad for toasters, but bad for nuclear reactors and data. Some manufacturers will report this number for a specific drive, some will not. If a manufacturer says a drive has an annualized failure rate of .9%, that means that at least 9 out of every 1000 of that drive they sell will die in the first year due to mechanical failure (this doesn't include other types of damage). This number is also influenced by RMAs.

Both ways of measuring failure rate are subjective and slightly flawed (as not all devices can be monitored, any drive not returned is assumed to be functioning properly). The failure rate of drives is not really valid after the warranty period either. These drives cannot be returned, thus any ability to monitor the product lifespan after this point is difficult to impossible. Thus, it is possible that the failure rate increases geometricaly, exponentially, or at a linearly increasing rate after this period (not like the day after, but you get the idea). This would probably have more to do with the failure rate of the individual parts, rather than the drive as a whole.

[hafa]

The drive reliability survey at Storagereview.com (http://www.storagereview.com/) has some interesting, albeit apocryphal data about drive reliability.

[WiNd]

that really like say the tranny on your car would die sooner from the wear and tear of shifting gears from drive to park all the time. Which can be true but its takes time, either way you'll find out that you need to replace your drive before the motor would wear out.