Monitoring HP-Procurve

These definitions worked like a charm on my 2910s.

For starters, the config given is very helpful (who the hell would know all those OIDs or dig through all the MIB files you can download?), but it can have unexpected results in case your switch model has more/other sensores than the ones used for reference here:

The commands given queries several sensors with fixed OIDs (hpicfSensorStatus.).
In most cases, the sensor index queried will actually represent the sensor wanted, but there is no guarantee for this.
Better to first query the list of sensor descriptions (hpicfSensorDescr, OID .1.3.6.1.4.1.11.2.14.11.1.2.6.1.7) e.g. by snmpwalk:
snmpwalk -c -v 2c enterprises.11.2.14.11.1.2.6.1.7

.1.3.6.1.4.1.11.2.14.11.1.2.6.1.7.1 => description of first sensor, .1.3.6.1.4.1.11.2.14.11.1.2.6.1.4.1 => state of first sensor;
.1.3.6.1.4.1.11.2.14.11.1.2.6.1.7.2 => description of second sensor, .1.3.6.1.4.1.11.2.14.11.1.2.6.1.4.2 => state of second sensor;
and so on.
That way you can be sure yor HP device does not in fact have any other mapping.

Additionally, you might want to know the possible stated and their meanings:
1 => unknown,
2 => bad,
3 => warning
4 => good,
5 => notPresent