check_hw_sensors (Visit this link)

Works like sensorsd(8) but reports to Nagios.  Allows you to monitor the hardware sensors that OpenBSD supports.  Things like fan speed, temperature and many more.

Examples of some sensors that can be monitored:

hw.sensors.esm0.temp0=32.00 degC (CPU 1)
hw.sensors.esm0.fan2=3667 RPM (Backplane Fan 1), OK
hw.sensors.esm0.fan8=0 RPM (Fan 1), CRITICAL
hw.sensors.esm0.indicator0=Off (Chassis Intrusion)
hw.sensors.esm0.drive0=online (Drive 0)
hw.sensors.esm0.drive2=unknown (Drive 2)
hw.sensors.safte0.temp0=27.78 degC, OK
hw.sensors.safte0.indicator0=On (Fan0), OK
 

This can be handled with just running:

check_hw_sensors

but that will report that esm0.fan8 is critical and OK(3)

Once you figure out that fan doesn't actually exist, you probably need a config file:

hw.sensors.esm0.temp0:warn.low=25:warn.high=35:crit.low=20:crit.high=40
hw.sensors.esm0.fan8:ignore
hw.sensors.esm0.indicator0:crit=Off
hw.sensors.esm0.drive0:crit=online
hw.sensors.esm0.drive2:crit=unknown

The sensors that report their status (OK, CRITICAL at the end of the line) are handled automatically.  Sensors that don't report that you have to give acceptable values.  The config file lists acceptable values, not values that are actually critical. So, with the above config as '/etc/sensorsd.conf' running:

check_hw_sensors -f

will tell you OK(7) or, possibly OK(55) or even more depending on hardware.

v1.42 now obeys the timeout from utils.pm