Jump to content
RESET Forums (homeservershow.com)

How much does operating temperature affect the failure rates of disk drives?


schoondoggy
 Share

Recommended Posts

My takeaway from that study is that they really didn't do a temperature comparison. If you look at the temperatures, they barely got above 30C, and that was only on 1 model. That's not running warm -- 50C is warm, 60C is really warm, 70C is hot.

 

I am totally NOT surprised they couldn't find any correlation, not at those temps. In fact, I would be amazed if they did find one.

 

As schoondoggy said, it would be interesting to see what happens when the drives are heated above operating temp. I would rephrase that: it would be interesting to see what happens when the operating temp of the drives is above 40C. Let's see what happens to failure rates when the average temps are between 40C and 90C; let's see if the failure rate goes up with the temp in that scenario.

 

I have pulled drives out of NAS racks that were so hot I could barely hold them, and these racks were in a conditioned server room with tons of air circulation. The problem, IMHO, is that the drives were just packed together too closely, with too little air circulation inside the NAS cage itself.

 

I have a feeling that one of the main issues with these studies is that they're measuring different things, so they come out with different conclusions.

Link to comment
Share on other sites

 Let's see what happens to failure rates when the average temps are between 40C and 90C;

 

Reminds me of the joke about the scientist and the frog. Scientist claps and yells "Jump!"; frog jumps. Scientist amputates one frog leg, then scientist claps and yells "Jump!"; frog jumps. Scientist amputates another frog leg, then scientist claps and yells "Jump!"; frog jumps. Scientist amputates third frog leg, then scientist claps and yells "Jump!"; frog jumps. Scientist amputates final frog leg, then scientist claps and yells "Jump!"; nothing. Scientist claps and yells "Jump!" again - still nothing. Scientist does this five more times with no movement from frog.

 

Scientist writes paper that frogs hear through their legs, and when legs are removed, frogs are deaf.

 

Or, more succinctly, correlation<>causation.  However, I think it's fair to say the vast majority of hard drives, especially those in the consumer market remain between the 30-50 C range so there is some relevance to the home use case, and likely even the home server society.

  • Like 1
Link to comment
Share on other sites

I know that frog joke; heard it at least 50 years ago :)

 

You make a good point, although I would argue that anything over 40C is undesirable with hard drives. I prefer to keep mine below 35C. Also, I've felt plenty of hard drives in desktop computers that I know were well above 40C. I vividly recall a specific example. Dell put out a whole series of their GX space saving chassis that practically cooked the drives. They got so many service calls they changed the design on later models. They added an extra cooling fan right below the HDD and I have to say it did a 1st class job of cooling the drive. It was actually a little surprising how well it worked.

 

I think my biggest takeaway from that article is that the BackBlaze chassis must be a superb design. They keep most drives between 20C and 30C. That's remarkable, and my hat's off to them. If only other manufacturers paid that much attention to their chassis design. Dave or pcdoc should have them on a show sometime.

Link to comment
Share on other sites

I have often wondered if living in a part of the world with lower ambient would allow all components to last longer. Here in Brisbane the Summer ambient is around 30C (86F) for months and never gets much below 20C (68F) in winter. StableBit Scanner shows all server drives running between 35 & 42 except for the SSD at 31. Ambient in the room is 26. Humidity is always high too which I feel also reduces the life of electronic components.

Link to comment
Share on other sites

  • 2 weeks later...

Higher humidity is better than lower, IMO. Little or no humidity means static electricity, and that's horrible for your system.  It's one of the down sides to San Diego. :(

 

And the data from BackBlaze should be taken as anecdotal. How many of those drives are shelled externals from around the flood time? And how many were "flood drives" in general. Etc. The evidence *is* very anecdotal. 

 

That said, more cloud companies should post statistics like this. Would be nice to see real-world statistics about drives. 

Link to comment
Share on other sites

Humidity high enough to prevent static discharges is highly desirable. Just don't overdo it. One of the issues encountered when going to the tropics (read rain forests) is that the humidity can be so high that algae and such will actually start growing on circuit boards, and at an extremely alarming rate.

Link to comment
Share on other sites

lol, true, But even slightly increased humidity is better that what is "natural" out here. :)

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...