Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

Have you ever had to alter your deployment plans because of thermal dissipation issues with your chosen hard drives? Was your solution to increase the cooling solution, or to move to a cooler drive solution?

share|improve this question
add comment

4 Answers

up vote 2 down vote accepted

More cooling. Depending on the circumstances, I did one or more of the following:

  • Adding a hard drive cooler to the hard drive directly
  • Separate hard drives if there was room to do so when there was more than one drive in a case
  • Add case fans taking into account case airflow - same number of fans blowing in as blowing out, and placed so the airflow would be even through the case
  • Added ventilation to the cabinet/closet/rack/etc
  • RAID mirror to lower down time and risk

It's almost always cheaper to properly ventilate the enclosure(s) than it is to pay extra for the cooler drive (which, in general, has lower performance anyway). Especially when you know that the drive may not last regardless, but the cooling solution will still be around for the next drive, so the overall cost is less.

-Adam

share|improve this answer
add comment

Probably not exactly what you're looking for, but I had a beefed up workstation that I used as a dev server about 7 years ago. I put an Adaptec SCSI and 4 9GB ultrawides in it. I tried adding about 3 different types of case fans that pulled in air from different places and that didn't even do it. Eventually just took the cover off the case. Made a great space heater under my desk.

share|improve this answer
add comment

We had a problem with the first generation of 2.5 inch SAS SFF disks from HP, their first-month failure-rate was WAY above the expected bell-curve, turns out they were too hot and they had to be replaced (over 3,600 of them too!) - that said they're rock solid now and great for what we need.

share|improve this answer
add comment

Probably best to do a combination of both . . . you can beef up the HVAC of the room but just like with hard drives . . . it's going to fail eventually and you need the hard drives to work in hotter that normal conditions just for those times. In data centers they are always tweaking the HVAC because as they add servers they increase the heat and at a certain threshold they'll add another AC unit.

The conventional wisdom that it needs to be 60 F (or whatever) in a server room so the hard drives don't fail is false . . . studies (by Google and others) have shown that increased temps in server rooms didn't make a significant difference in hard drive failure rates (up to a ceiling of maybe 80 F or something like that).

I would get the "better" hard drives that run cooler assuming the cost difference isn't too crazy but even then it could be too hot so you have to get the HVAC under control too.

share|improve this answer
    
Actually, I was thinking less about the issue of room cooling and more about the issue of keeping your specific servers in their enclosures cool. But interesting response nonetheless. –  Paul Sonier May 7 '09 at 21:13
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.