By monitoring devices (e.g., printers) with SNMP, we can get device uptime measured in "timeticks". But who is interested in the working time measured in hundredths of a second (because this is how the unit of time measurement is defined in this case)? In general, we would prefer this parameter given in hours.
Operations Manager can scale performance counter values on charts using powers of 10.
However, to achieve the desired figure, we need the value divided by 360000. Since the graphical interface of performance charts does not do it, you can transform the collected data.
If you look into the XML form of the rule collecting performance data, you will see a fragment:
The calculated value would have to be divided by a factor of 360000 for the measured value to be expressed in hours, not hundredths of a second.
The performance counter value, specified as ComputedPerformanceValue will come from dividing (Division) two numeric values (NumericValue). The first of these values, is a reference to the same SNMP element as in the original expression, while the second expression is in the form of a number by which the division is carried out. In this way, we get:
Finally, the operating time (Uptime) in hours is shown:
A similar approach can also be used if one OID stores a counter value and another stores a unit of measure. In the latter case, it is a static value that should come from the Discovery process, not from the monitoring itself.