You are on page 1of 4

Each HP model has the temperature operating range described in the config guides and data

sheets.

There are many ways to check and/or create alarms based on ambient changes like temperature.

Note that temperatures are measured on airflow IN to the system rather than airflow LEAVING
the system.

In practice, you do not need to know the temperature, since the EMS warnings are triggered when
action is needed to prevent the system shutting down. If you are getting the notifications, then you
need to do something about cooling. The hardware knows what are acceptable temperatures (it ma
y vary between server types), and triggers the events when limits are reached.

In any case, measuring the temperature depends on the firmware and hardware. I enclose here wi
th some of the possibilities to check the current temperature values:
a) cstm can do it in some versions.

b) EMS will raise alerts whenever an exceptional condition is met. That can be captured by OVO,
and many other tools, including open-source Munin-node and Nagios

c) SFM has Temperature Sensor Provider.


The Temperature Sensor Provider retrieves the properties such as sensor number, current tempera
ture reading, and temperature sensor status.

An example:
# sfmconfig -a -l -t SystemTemp -d NULL
Caption DeviceID Status
HP_SystemTemperatureCollection NULL OK

d) Access via console can provide some details (how much - it depends on the model and firmw
are). Here is one from rx2600:

MP:CM> ps
System Power state: On
Temperature : Normal

From blade server BL860c:


For System Processor Status see the SS command
For Power Supply and Fan status see the Onboard Administrator.
System Power state : On
System Power usage : 203 Watts
System Power allocation : 255 Watts
Ambient temperature : 24 C
Temperature status : Normal
Over Temperature protection: Enabled
e) Some models, have another option through console. Here are the results on rx7620:
MP:CM> de
Display status of the selected MP entity (for use by trained personnel only)
B - BPS (Bulk Power Supplies)
U - CLU (Cabinet Utilities: Fans, Intrusion, Clock's etc.)
A - PACI (Partition Console Interface)
G - MP (Management Processor)
P - PM (Power Management)
H - Cell Board Controller (PDHC)
Select device: u
Cabinet 0 Utilities Status
FW Revision : 8.006 built Mar 23 2007 at 16:24:46

PWR SBY MP RUN REM ATT FLT


Front Panel LED State : * * * * * * .

Inlet Air Temperature : 20 deg C


Altimeter Status : OK
Operating Altitude : Below 1000 ft (304m)

Fan Rate ( rpm/30 ) : Standby Main I/O


97 97 90
Fan Status :
012345
Stby/Main : N N N N
IO : N N N N N N
(N:Normal Operation -:Unpowered f:Failure Imminent F:Failed)

f) If you like System Management Homepage(SMH), than this works nicely:


Home -> System -> Temperature
Status Sensor Location Temp Threshold
OK Temperature Sensor 216 Ambient Temp 22 38
OK Temperature Sensor 217 Processor 0 Temp 62 103

g) On HP-UX 11.31, the cprop(1M) command provides a command-line interface to most of


the information available from the System Management Homepage property pages.

Here is a result from HP Education HP-11.31 server running March 2010 Update:
# cprop -summary -c "Temperature"
[Component]: Temperature
[Table]: Temperature
-------------------------------------------------------
[Instance]: 1
****************************************************
[Hash ID]: Temperature:2796559075
[Status]: OK
[Sensor]: TempSensorInfo 1: Proc 0 ThermTrip
[Location]: CPU board
[Temp]:
[Threshold]:
****************************************************
[Instance]: 2
****************************************************
[Hash ID]: Temperature:2740717654
[Status]: OK
[Sensor]: TempSensorInfo 2: Proc 1 ThermTrip
[Location]: CPU board
[Temp]:
[Threshold]:
****************************************************

h) Another open-source tool can be used -


ipmitool. The source code is available from the Sourceforge site:
http://ipmitool.sourceforge.net/
It is easy to build it from the source code.

If "IPMI over LAN" is enabled in the MP configurations, you can do the following:
# ipmitool -I lan -H -P "" sdr type Temperature
Ambient Temp | D8h | ok | 23.1| 16 degrees C
Processor 0 Temp | D9h | ok | 3.1 | 53 degrees C
Processor 1 Temp | DAh | ns | 3.2 | Disabled

Finally, take a look at envd(1M). The envd daemon provides a means for the system to respond to
environmental conditions detected by hardware. Such responses are typically designed to maintain
file system integrity and prevent data loss.

envd works with two threshold levels for environmental temperature: critical and emergency
(OVERTEMP_EMERG and OVERTEMP_CRIT). By default, when emergency threshold is reach
ed, envd issues a shutdown.

In general terms:
OVERTEMP_CRIT = over 30 degrees Celsius
OVERTEMP_EMERG = over 34 degrees Celsius -
the usual response to this is for envd to complete a graceful shutdown

There is another event at about 40 degrees Celsius which will cause the platform monitor to just
remove DC power to the system (non graceful halt), but unless temperatures go up VERY quickly
you should not reach that one.
You cannot change any of these values nor should you even attempt to - HP has determined the
safe operating envelopes for the systems, changing them would probably invalidate warranty.

You might also like