What is the best way to read temperature data from a blade, specifically an HS22 or QS22 in a BladeCenter H chassis?
As far as I can tell, the AMM only displays the media tray temperature. My OS is Red Hat Enterprise Linux 5 so I originally tried to use the lm_sensors package but it was unable to find any sensors on the HS22. Is there a certain kernel module that I need to read the sensor data? Alternatively, does IBM or Intel offer a program to read this data?
Any help would be greatly appreciated. Thanks in advance.
Re: HS22 and QS22 temperature monitoring2010-01-28T15:04:48ZThis is the accepted answer. This is the accepted answer.I think you must install the IPMI Stack and IPMI tools to query the sensors.
Just a few links which might help or give an idea - since nobody else answered.
IPMI on Linux
Using Intelligent Platform Management Interface (IPMI) on IBM System x Linux Platforms
IPMItool - Tool to connect to an IPMI supported machine
An example from a IBM x ( HMC ) linux system.
$ ipmitool -U USERID -P PASSW0RD -H 10.1.0.2 sdr
Fan 3 Tach | 1350 RPM | ok
Planar 1.5V | 1.56 Volts | ok
Ambient Temp | 35 degrees C | ok
CPU OverTemp | 0x00 | ok
CPU PFA | Not Readable | ns
Re: HS22 and QS22 temperature monitoring2010-01-28T17:00:09ZThis is the accepted answer. This is the accepted answer.Thanks! That looks promising. Unfortunately, I won't be able to look into that more until Monday. I'll post back here with my results though.
Since making my first post I was able to use lm_sensors with the coretemp module loaded into the kernel to read 16 CPU sensors from the HS22. Our HS22 has 2x quad-core Xeons with HT. I know most OSes will read that as 16 CPUs, but I'm was really only expecting 8 sensors (one per core). The only problem with using lm_sensors is that the latest kernel for RHEL/CentOS 5.4 doesn't support the coretemp module. I was doing all of my testing from an Ubuntu 9.10 x64 LiveCD. I'd love to be able to get temperature sensing from RHEL/CentOS.
Re: HS22 and QS22 temperature monitoring2010-02-03T15:19:39ZThis is the accepted answer. This is the accepted answer.OK, so I got the IPMI driver loaded and the IPMI-tool running but I don't see any temperatures.
I see some lines like the ones listed below:
CPU 1 OverTemp | 0x01 | OK
DIMM 1 Temp | 0x01 | OK
Re: HS22 and QS22 temperature monitoring2010-02-03T16:01:54ZThis is the accepted answer. This is the accepted answer.
- Eric K 270002WH82
But from the AMM i can see - sorry not checked before.
system> env -T system:blade
Component Value Warning Shutdown Reset
CPU 1 Temp 40.00 94.00 125.00 89.00
CPU 2 Temp 36.00 94.00 125.00 89.00
In case you only need the temperture i would go the AMM way and skip IPMI since this requries the IPMI stack to be installed where the AMM way is OS independend
Just found a link:
Using Intelligent Platform Management Interface (IPMI) on IBM Linux Platforms
Re: HS22 and QS22 temperature monitoring2010-02-03T21:02:52ZThis is the accepted answer. This is the accepted answer.Thanks for all your help so far, Hajo, but I'm not quite finished yet.
Using the AMM to read the sensor data will be sufficient for us and should be simpler. I tried it out and it read temperatures for our QS22s but not for our HS22. When I run the temps command it says "No temperature available for blade" I've tried this with two different HS22s now with the same result.
Also thank you for taking the time to find that other link for me, but it is simply an online version of the pdf you had linked in your first post. I've been using that first pdf quite a bit.
Re: HS22 and QS22 temperature monitoring2010-02-03T23:59:50ZThis is the accepted answer. This is the accepted answer.
- Eric K 270002WH82
Ups - i am very sorry - we have JS12 and HS22 blades and i tested the JS12 and not the HS22
An output from our HS22 gives:
No temperature available for blade
I just checked the Service Guide for the HS22 and it mention led for displaying errors.
A quick test
$ system:blade> led -l
Component Label State Location
Blade4 OverTemp Off System board
so you can check at least the state. In case an error arise the AMM should have more information - Error code is somewhere in the service guide
Re: HS22 and QS22 temperature monitoring2010-02-04T16:24:03ZThis is the accepted answer. This is the accepted answer.
- HajoEhlers 0100001U0A
Thanks for all your help, and if you think of anything else, please let me know.