Does Hard disk drives + heat = death?

[fffblackmage]

I was just wondering what other people's opinion on hard disk drives and heat are. Most people seem to generally accept the idea that HDDs die faster if they run hotter.

I should probably note that I believe temperatures outside normal operating temps. are unsafe. Therefore, it should be given that any HDD would die pretty fast in such unreasonable environment.

Do you buy things that are designed to keep HDDs cool? Or do you simply rely on fan(s)? Or maybe you don't have any kind of direct active cooling.

This report is a bit old, but maybe some of you haven't read this.

Failure Trends in a Large Disk Drive Population

Does that change your thoughts on HDD cooling?

------
My 900 has fans that keep the HDDs cool anyways, but my cheapy HTPC has no dedicated fan for keeping the HDD cool. It relies on the slow and quiet rear exhaust fan to slowly pull some cool air front the front of the case to cool the HDD. Hasn't died yet and it's an old WD 160GB HDD that I was using in my old Athlon XP computer.

My aunt also has an old computer that didn't have a dedicated fan in the front of the case keeping the HDD cool. It just relied on the psu and rear exhaust fan for moving hot air out of the case. The HDD lasted about 7 years or so before dying a month or two ago. The comp specs are Athlon XP 2000+ on a MSI VIA-chipset mobo with a GeForce 2 MX200, so that should tell you that the comp was quite old. The deceased HDD was a WD 80GB.

One other time I had a HDD die was probably caused by an ESD. It sat in an Aluminum external enclosure, and one day while it was on, I somehow managed to give it an ESD. The LED on the front stopped working, and soon after, other problems started showing up. It would disconnect randomly while still physically connected and on running, then later, it started making clicking noises of death.

 

[Raduque]

I never had a CPU that I could attribute it's death directly to heat. There was always some other factor (shock, bad power, bad cable, bad controller) that was more likely to be the cause of death.

However, I did have an older WD Caviar 30gb drive that had an extremely large amount of bad sectors. It lived stacked on top of a Maxtor 30gb drive for 3 years.

 

[Brian Stirling]

I have had about 30 HD's over the last 25 years and of them I've had about 6-8 HD's die on me. The average drive had been in service for 3 years or less and the longest a drive was in service was about 6 years.

Of the drive failures I would say that only one died an early death and can be attributed to manufacturing defect. The rest, in my view, died because of heat. As an example, I traveled the country in my RV back in 2004 and used a Dell laptop running Delorme Street Atlas and a Garmin 12XL GPS unit to provide moving map routing and navigation. The Dell laptop was a little over two years old and never had a problem, but only a couple weeks into the trip the laptop died and I later pinned it down to HD failure. I then realized that the laptop was in direct sunlight quite a bit and that it was getting hotter than normal. I modified my setup to protect the laptop from the Sun and had no other HD's problems for the remaining 10 months of the trip.

In the other HD failure cases I seem to recall most of them were the drive that got the least cooling and operated the hottest.

So, I would have to concur with most of the previous studies that suggest heat as the single biggest cause for HD failures...


Brian

 

[realmike15]

I have had about 30 HD's over the last 25 years and of them I've had about 6-8 HD's die on me. The average drive had been in service for 3 years or less and the longest a drive was in service was about 6 years.

Of the drive failures I would say that only one died an early death and can be attributed to manufacturing defect. The rest, in my view, died because of heat. As an example, I traveled the country in my RV back in 2004 and used a Dell laptop running Delorme Street Atlas and a Garmin 12XL GPS unit to provide moving map routing and navigation. The Dell laptop was a little over two years old and never had a problem, but only a couple weeks into the trip the laptop died and I later pinned it down to HD failure. I then realized that the laptop was in direct sunlight quite a bit and that it was getting hotter than normal. I modified my setup to protect the laptop from the Sun and had no other HD's problems for the remaining 10 months of the trip.

In the other HD failure cases I seem to recall most of them were the drive that got the least cooling and operated the hottest.

So, I would have to concur with most of the previous studies that suggest heat as the single biggest cause for HD failures...


Brian

I've had a much better experience. I have an 80GB Maxtor that's probably closing in on 8-10 years that still runs. My 60GB Maxtor which was bought a year or two after it also still runs. My 120GB Maxtor continues to run, which is probably close to 5 years old now. Bothy my 250GB Western Digital's which are about 3+ years old run great. If I remember correctly they are all 7,200 RPM drives.

I don't want to debate on what the most common cause of HDD failure is. But I can say if your drive doesn't fail within the first 3-6 months... it's very unlikely to fail for a very long time.

The question is how hot is the drive running, have you actually run a program to get a read out of the temps. If it's only slightly over normal temperatures, you're fine. One of my 250GB WD's has been running at 55 Celcius for 3 years and I've never had a problem or seen any signs of a wearing drive. Now if your drive is running at say... 70 celius that is something to be concerned about effecting it's lifespan. Also remember the faster the drive the more heat it produces. A 5,400 will generally run cooler than a 7,200 which will run cooler than a 10,000. I swear by Western Digital for reliability.

 

[Blain]

Modern perpendicular recording HD's = Higher performance, but quicker death than their older longitudinal brethren.

Such is life... ^_^

 

[Yellowbeard]

This is pretty much true of any electronic component. Of course, the rate of acceleration is going to increase the closer you get to the maximum operating temperature of the device.

IIRC, rotating HDs are rated to operate in the (appr) 60c - 100c range maximum. So, if you were comparing longevity for a drive at 70c vs the same drive at 80c it may take quite a while to see any difference.

 

[VirtualLarry]

I had a 75GXP (Deathstar) 30GB HD. The only time I got some bad sectors, was when it was in a mobile rack, without a fan, and it overheated. You could hear it have problems when it got above about 55C.
I try to keep my HDs at 40C or below now.

 

[Voo]

Well I don't see any reasons why the google study should be noncredible. Afaik google buys cheap disks that didn't pass all the QA tests from manufacterers, so their failure rate is probably quite a bit higher, but other than that they have a gigantic sample.

I'd believe them much more than single failure reports, after all there are many reasons why a HDD can die and there are always black sheeps. Also while traveling around with a laptop I'd say chances are good that the disk got some bad hits while running..


For me I just try to keep the disks in some reasonable temperature range (<45&#176 and I'm fine with that.. but who knows, maybe I'm just lucky.

 

[Russwinters]

It's possible; although most of the failures are going to be circuit board related, and very fixable/recoverable.

 

[Blain]

Has anyone actually scanned past the pretty graphs and actually read the "Conclusions"?


_F1_ for Yes
_F2_ for No
_F3_ for What?

 

[Voo]

Has anyone actually scanned past the pretty graphs and actually read the "Conclusions"?


_F1_ for Yes
_F2_ for No
_F3_ for What?

Actually I read the whole thing when it came out.

I'm not exactly sure what your point is, but it seems like you want to tell us that we should read the whole thing, because temperature has a influence on the disks, if that's so I must've read something differently.. "[..]reported temperature effects only for the high end of our temperature range and especially for older drives. In the lower and middle temperature ranges, higher temperatures are not associated with higher failure rates"

Which is more or less exactly what at least VirtualLarry and I said and probably most others here meant.
Though I'm not sure how you should even reach 45° or something like that in a common PC, my two HDDs here have 27/35° atm and there's no mentionable air flow around them.

 

[RebateMonger]

I'm pretty certain that I've "melted" at least one hard disk. In 2003, I had a brand-new Maxtor that I stacked into a plastic enclosure surrounded by two other hard disks and almost zero ventilation. I then copied 200 GB of data to the the disk. About five hours into the copy, the disk failed. When I pulled it out, it was too hot to touch. It remained dead after cooling.

Maxtors of that generation had known issues with overheating their PCBs, resulting in disk failure. So it wasn't the best idea to squeeze it into a tight space and copy 200 Gigabytes of data to it all at once.

 

[Emulex]

google runs their datacenter in the 80's - its really how efficient you move the air - study your fluid dynamics.