Rising System Temperatures

by Bob May 9, 2008 09:22 PM

Posted in: HP MediaSmart Server

I purchased an HP EX470 about 2 month ago from Newegg and have been very happy with it until a month ago.  I have started to get a ramdom messeage that my system was at risk.  Opening the Hardware Staus window I discoverd that my system temperature was "overheating" with a temperature of 61 to 62 degrees C.  It had been running around 29 to 31 degrees C.  The MSS box itself was still cool to the touch, both fans were running 1,418 to 1,430 RPM, and the processor temperture was running 27 to 29 degrees C.

I called the HP Support line and they told me this was a "known" problem that is either software or hardware related. I was instructed to do a factory reset.  If the problem went away for a while, then it was a software problem and a patch would be out with the PP1 upgrade.  If it came right back, then it was a bad sensor.  We'll my luck was not running that day since the system warning came right back.  

I called HP support with the results of the test. They are now stating it is only a software issue and they were wrong to tell me that the sensor was going bad.  So, I am NOT sending my system back at this time. Crying or Very sad

I read in another forum that I could change the temperature warning limit my editing the "HP MediaSmart Server.ini" file.  I did that by changing the temperature warning to 75 degress C but my system temperature has continued to increase.  It is now at 70 degrees C and rising one degree every few days.  The box itself is cool to the touch and the fans & CPU are running normally.

What I would like to know is this really a software issue that is going to be fixed with PP1 or do I need to keep bugging the Support people about a replacement.

Thanks.... 

Discussion:

Rwkessler on May 11, 2008 2:22 PM

Sorry to hear that you too have this issue.  I reported this same issue to HP 2 weeks after starting up my HS.  I was first told to do the server recovery, that it should correct the incorrect reporting of the temperature.  I called back a week later just to make sure and was told that the error would come right back and to not bother do the server recovery.  I was given a case number and told that HP was working on a software fix and would be sending it out as an update.   This last communication was mid April.   Like you, my HS temperatures have risen, today they are:

Upper Fan: 1493 RPM    System Temp 86 degrees C or 186.8 degrees F

Lower Fan: 1480 RPM   CPU Temp 26 degrees C or 78.8 degrees F

 I know that the system temp is wrong, the temperature in the top of the server is lower than the temp coming out at the bottom where the CPU resides.  I only have the one drive installed so there is no heat source at the top of the box to drive the temp up.

I too, do not like that my system is reporting such high temps in error and certainly hope HP comes out with the software fix soon. 

Greybeard on May 27, 2008 9:10 PM

I had the same problems within 1.5 weeks of receiving my WHS. Initially I was told by tech support that it was probably a software issue and to wait for an
update the following Tuesday (20 May). On 19 May I got a message from tech support telling me that it appeared to be a hardware problem. Sensor reported 80° celcius. I was instructed to return to HP and they would ship a replacement
when they received mine. Contactedcustomer service with the case# from tech support, got my Return Merchandise Authorization and shipped it back on Tuesday. Received replacement on Friday.

JohnBick on May 28, 2008 7:03 PM

I suggest that you get the SpeedFan application and install it (via RDP) on your MSS. One thing you will immediately discover is that the MSS fans run at only 50% speed! I would recommend increasing to 100% and watching the temperatures. (Use the "Chart" tab in SpeedFan.)

I have set up mine to automatically adjust the fan speeds to control the temperatures AND to set the speed to 100% on each fan if the application is closed. My MSS runs cooler this way and, I hope, this will result in a longer life.

Temperatures WILL increase as the CPU is loaded -- and it happens fairly quickly, too. So if you have added Add-Ins or applications that use CPU cycles this will raise the internal temperatures. You may want to check the process activity (use Task Manager) to see if there is something loading the CPU constantly. (DEMIGRATOR.EXE is the "balancing" function -- that can really run a long time if you have the "stock" CPU and memory configuration -- I know of one case where that was found to be the cause of a rising temperature. (At least one other case very similar to yours was found to be caused by a constantly running application. When disabled the temperature returned to normal!)

Be careful comparing the SpeedFan temperatures to the System and CPU temperatures reported in Settings-->Hardware Status nto make sure you are comparing the right ones. Also note that the Temp1-3 indicators do NOT seem to be the disk temperatures. Also note that you may at times see six temperatures reported and, at many (most?) times only five. We haven't figured out why the intermittent one (a number of folks report it as intermittent) or where some of those sensors are located.

Speaking of disks: If you use the Disk Management Add-In it will report the actual disk temperatures based on a WMI query to the drives. The Client Info Add-In does the same thing -- but the temperature reported is slightly (but consistently) different. Seagate drives tend to run 30-37C and the WD 1TB drives run 23-32C (using Disk Management).

If all of your temperatures run hot using all these packages I would not think it is a software problem.

A good place to continue this discussion would be over on MediaSmartServer.net. I started a thread on SpeedFan there to consolidate discussions on that application:

http://www.mediasmartserver.net/forums/viewtopic.php?t=1236&highlight=speedfan

There are also a lot of other "temperature" discussions in other threads, and a good "search" tool to find them!

JohnBick on July 15, 2008 7:44 PM

Out of curiosity, have you tried re-seating the backplane? (Disassemble as though about to add RAM, remove the backplane, replace it, reassemble.) This has worked for several folks who incurred temperature warnings immediately after they did upgrades.

The sensor is at the top of the backplane and the A/D converter is on the motherboard. There is an analog signal being passed down the backplane, across the connector and onto the motherboard. It doesn't take much to have a "dirty contact" and the "wiping" action of removing/replacing may clean the contacts sufficiently for you to get a good connection.