[Techtalk] Temperatures in server rooms

Magni Onsoien magnio+lc-techtalk at pvv.ntnu.no
Thu Apr 19 12:26:49 UTC 2012


Traditionally the recommended temperatures in server rooms have been
20-25C. Recently it's become a few degrees higher, since this apparently
doesn't affect the life of servers, and since the cost/benefit
calculations have changed with the increased need for power, increased
prices and increases temperatures (and other similar factors not
relevant here).

Google published a paper on the lifetime of consumer-level SATA drives
(http://research.google.com/archive/disk_failures.pdf) which basically
says that HDDs will survive higher temperatures than previously
expected, and also that the lifetime of the disk decreases at rather low
temperatures and increases at high temperatures (40C). It is important
to note that their temperatures are measured by reading SMART-data from
the HDDs, i.e. are the temperatures on the HDDS, inside the cabinet, not
the temperature of the air, which is what we'd usually measure when
talking about the 20-25C temp. So it's also important to know the
difference between the temperatures inside the cabinet and in the room
as such.

So, how does the room temperature matter? How does it affect other
components? Fans and power supplies spring to mind, since they have to 
spin faster/more with higher temperatures. And the consequences of 
failed fans can be quite...bad. I have not managed to find any data on 
this.

Anyone? Papers/good references would be most appreciated, but also data
points. Basically I find the current 30C (measured in the middle of the
room between the rack rows, so I guess it will also be the temperature
of the air drawn into the server) a bit too high. 
The room is used for production systems, but AFAIK not for critical
external services.


Magni :)
-- 
sash is very good for you.


More information about the Techtalk mailing list