I have a 3 - 4 year old 250GB WD hard drive that has shown abnormally high temperature readings ever since I bought it. The temperatures were so much higher than expected that I ended up attributing them to a bad sensor. For example, the other drive I have on the my system at the moment is a WD 640GB and it is currently at 40C. While the 250GB drive is at 62C. Typically, when the system boots up the temperature for that drive will quickly clime to 50C - 58C (today's a bit hotter). These drives are right next to each other in the case so that there is no other source of heat that could affect the temperatures.
All in all I never had any problems with that drive (except for the alarming temp readings) but in a recent S.M.A.R.T data check I noticed some CRC errors. These errors are usually attributed to bad cables bad just to be on the safe side I am planning on retiring this drive soon.
In the early days, when the 250GB drive was my main drive I was quite worried so I bought a hard drive cooler for that drive. It manages to to lower the temperature by about 5C. The problem is that it really makes a lot of noise. What I am interested in knowing is if my assumption that it's more than likely a bad sensor is logical so that I can disconnect the hard drive cooler and get some peace and quiet.
if it is the drive i am thinking about, a wd 250gb jb series, then speedfan does give the wrong temp reading, but if it is about 4 years old then i would be very careful of it since i just had the one i was monitoring go with crc errors. right away we swapped to other drives and the drive became unreadable later in the day. be careful and migrate away from that western digital drive if you have critical data on it. if you want the model of the drive i can look it up remotely on one of the old servers and post the model back here.
It sounds like this isn't a rack mount server but a tower type of system.
If so, and you don't have high uptime requirements, try removing the "hot" drive from the usual area next to the other hard drive and running it there. For example, if you have something like a small cardboard box it can rest on rather than mounting in the case. This is temporary while you test the thermal issues.
There are two things you can test from this.
Does the temperature run more similar to the other drive now?
Does the drive feel hotter to the hand than the other drive while touching the metal casing?
If the temperature is reduced a great deal, then the problem can be because it was sandwiched next to the other drive. Generally, heat will rise and the top drive will be hotter. Drives will also run cooler when not sandwiched together too closely. I prefer to leave 1 inch of air around if possible.
If the temperature feels the same to the hand, but the probed value is more than 10 degrees difference, maybe the sensor is inaccurate.
The temperature isn't innacurate by a "large margin", in my opinion. I would say that the reading of ~60 degrees is accurate. While it MAY be possible that the sensor has an offset, I'd say that it's probably just a model that happens to run abnormally hot. Normally, temperatures above 50 deg C aren't good, and coupled with CRC errors and the age of the drive, you'd be best off replacing it (and make sure you have a backup!).
Yes, that is possible.
I've had SSD-drives that are measured as being -20 Celcius. Which was really bad, because ACPI sendt a "power-button"-event to the operating system shortly after it was booted up.
In the morning, when starting a cold computer, the temperature in both drives should be the same( +-1 degree). If the diference is large, then one drive is showing a wrong temperature.