16

Background

My personal desktop system at home has 5 SATA drives racked up inside. Recently my system started failing in odd ways like random kernel panics and I eventually traced it to random degrades on the RAID array. Sometimes I could boot, other times I couldn't and so on. After chasing software issues for a while I finally went to pull the drives and discovered the real reason they were failing: they were hotter than a barbecue on the 4th of July! The front case fan had seized up and the PS fan had a loose power connector caught in its grate so the inside of the case had been cooking.

As a hold over, I found a house fan and got that sucker cooled off. It ran great with everything nice an chill. About this time I learned how to get drive temperature readings from S.M.A.R.T.

for i in a b c d e; do
    sudo smartctl --all /dev/sd$i | grep Temperature_Celsius
done

Now I know that with my case opened an a house fan permanently cleaning out the cobwebs the drives run at 31-32°. A quick test with no ventilation to replicate the failed state shows the drives ran up to the high 40s pretty quickly. I don't know how bad it was during the actual failure or how long its been like that.

With this in mind I replaced the failing fans, added a couple more, upgraded the front one blowing across the drives from 80mm to 120mm and closed it back up. With it standing back upright again the temp range is now generally sitting at 32° on the bottom of the set and 37° at the top.

The Question

What is a general safe operating temperature range for SATA drives? Should 37° be a concern or is drive damage not an issue until after a certain point?

Although the drives seem to test out fine now, how likely is past exposure to heat likely to make them prone to failure now?

2
  • One of the things I noticed especially with the newer tech cooler running drives, they get a LOT hotter when under hard use. While my greens (for example) are practically cold most of the time I am in there looking/touching, they still get warmed up good when they are working hard. Same thing with the Sata controller chips, Suuuure the Intel controller don't need the big passive sync I added to it, until I Probe it when it has actually been working hard. I am saying that tests (when looking) rarely represent reality when stuff went ary. Everything else was said, they can easily work 50*C
    – Psycogeek
    Dec 13, 2011 at 1:55
  • Related: serverfault.com/questions/25611/…
    – user11574
    Aug 27, 2018 at 23:25

4 Answers 4

26

37 degrees should not be a problem at all. Naturally, hard drives differ in their specs, some can run hotter than the others. You should check the published specifications of the drives that you have. For example, WD Caviar Black 1TB operational temperature is -0° C to 60° C. Of course, you would not want your drive to run 60°, as it might reduce its life span.

Google published a very interesting study (PDF) about hard drive health and lifespan, based on data collected from their systems (many thousands of hard drives). That study says that:

Overall our experiments can confirm previously 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. This is a fairly surprising result, which could indicate that datacenter or server designers have more freedom than previously thought when setting operating temperatures for equipment that contains disk drives.

Their graph shows that faulure rate does not go up until drive temperature goes past 45 degrees.

3

This is only my limited personal experience, but I have run a few drives in the upper 40C without issues over 2-3 years, since I was using a silent enclosure (cooled by a small fan to avoid reaching 50C). At these temps I would assume shorter lifespan and quick death, rather than random data corruption, but I could be wrong. In any case, anything under 40C-42C is just fine.

But don't underestimate other factors likely to play a role in data corruption:

  1. The south bridge that houses the IDE/RAID controller chipset is often cooled by a small heat sink only. They tend to run hot in normal conditions, so a rise in ambient temperature due to a lack of case airflow and lots of HDDs could plausibly cause data corruption.

  2. RAM or CPU overheating is a common culprit of CRC memory errors, which translates to data corruption. Monitoring CPU temp and performing memory tests is essential when faced with data corruption.

If your current S.M.A.R.T. indicators are fine and don't show uncorrectable sector counts, I would consider the drives safe for use.

0

The initial post stated that the user did not know how hot his drives had got - some discs record this parameter and it is accesible via the SMART information. Hard disc sentinal is one piece of saftware that reports this as maximum temperature in entire lifetime.

I have seen results for Maxstor, WD & Seagate drives

A couple of my external USB drives that I use for backup show maximum temperatures of 63 & 64 C respectivly ! I have now built a cooler for the external drive

0

I've read the report mentioned above and check operating temp ranges for samsung, WD, and hitachi hard drives. Based on research I have concluded that drives operating in the 30 degrees to 45 degrees offer the least likely hood of failure; A S.M.A.R.T. high temperature of up to 55 degrees is not a cause for any concern; and that Peek temperatures over 60 degrees would indicate a reduced drive life expectancy.

My own testing indicates that a rise of up to 10 degrees is to be expected during periods of heavy access.

There appears to be some unknown factor at work that is causing abnormal failure rates in external drives. While heat seems to be part of the problem it is not the whole answer. I would advice all users of external drives to monitor temperatures closely whenever there is a change in the nature of the drives usage or environment. This appears to be some "unknown" (not the usual) problem that leads to these drives over heating unexpectedly.

I am currently rejecting the explanation of computer virus, user ignorance, bad USB protocols and the placement of the drives in an improper location.

1
  • 2
    Welcome. The purpose of this site is to compile high quality questions and answers that people can rely on for reference. A couple of things in your answer may make readers wonder. You discuss research and testing. If this is published, it would be valuable to cite that. Your use of the word "unknown" raises a question about your qualifications. It would help to describe what known factors you ruled out and what "usual" problems cause over-heating. Describe your tests and how many drives you base your conclusion on so people can understand the scope.
    – fixer1234
    Nov 11, 2014 at 5:24

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .