Thermal Warning

Top  Previous  Next

This feature can be added to the command-line and run in the background as part of the scheduled polling process.


When you invoke smartmon-ux with the optional -G threshold_temperature warning, you instruct the software to not only monitor SMART alerts, but to also report alerts if the disk temperature meets or exceeds the supplied temperature.


This feature requires that you have a SCSI, Fibre Channel, or SSA disk that reports temperature via the ANSI-defined temperature log page entry (found on page 2Fh). If you are not sure whether or not your disk reports temperature here, you can just try the command and give it a threshold of 1 degree C. If your device supports temperature reporting, you will get the alert in the syntax reported below.


Temperature polling does not have any significant additional load and is a convenient fail-safe to insure your computer does not run too hot.


Note that you can also monitor temperature via sophisticated scripts, even if drive temperature is not reported on page 2F (hex), but is reported in a vendor-unique page that smartmon is already aware of. Use the threshold configuration and threshold monitoringfeatures to create temperature over time log files or warnings if temperature increases 5 degrees or more over a few minutes.


[root@rh90 smartmon]# ./smartmon-ux -F 600 -G 40-L  /dev/sda

SMARTMon-ux [Release 1.21, Build 26-JUL-2003] - Copyright 2003 SANtools, Inc.

Discovered SEAGATE ST373307LC S/N "3HZ0381E" on /dev/sda (Enabling SMART)(70007 MB)


Launching job #27401 in background - Will poll every 600 seconds.


This instructs the software to check temperature and report a thermal  warning if temperature exceeds 40 degrees C. (Temperature is always monitored and reported in degrees C, not degrees F).


The -L option instructs the software to log results in the file smartmon-ux.txt. In the case of LINUX, the file is saved in the /var/log directory.


Below is the tail end of the log file. Since the disk was at 42 degrees when the program was launched, the text highlighted in red will be added to the log file.


Sun Jul 20 20:24:03 2003: ./smartmon-ux started

Sun Jul 20 20:24:04 2003: Discovered SEAGATE ST373307LC S/N "3HZ0381E" on /dev/sda (Enabling SMART)(70007 MB)

Sun Jul 20 20:24:04 2003: /dev/sda polled at Sun Jul 27 20:24:04 2003 Status:Passed (Temperature = 43C 109F)

Sun Jul 20 20:24:04 2003: Device on /dev/sda, Thermal alert. Temperature now at 43C 55F degrees.


Once a thermal alert is sent out, they are not repeated every polling cycle. The temperature has to go below the threshold to reset the trigger to allow thermal alerts to be logged.  If you launched the program with -G 45, the line in red would not get added to the log file.


If you always just want to report the temperature, but do not want thermal alerts, pass it some high temperature, such as -G 99. As this is over 200 degrees F, your computer and disk drive would have shut down (or melted), long before that, and you will never get an alert.


If you want to know what temperature your disk will enter a thermal alert on its own, you can either read the disk drive's specification (which is difficult to find), or ask smartmon-ux ....


./smartmon-ux -C /dev/sda

SMARTMon-ux [Release 1.21, Build 26-JUL-2003] - Copyright 2003 SANtools, Inc.

Discovered SEAGATE ST373307LC S/N "3HZ0381E" on /dev/sda (SMART enabled)(70007 MB)

Statistical log pages dump below [# of bytes reserved for value in device]:


(.. truncated here)


  Current temperature +/- 3 degrees C: 41

  Reference temperature +/- 3 degrees C: 68


The current temperature is 41C, and the shutdown is 68C. You can see that this disk drive can run much hotter before there is need for concern.



The temperature for SCSI, SAS, and fibre channel devices is standardized, but optional.

You can also obtain device temperature via the log page viewer.