The monitoring node provides the ability to record a monitoring record. At run time, the value of msg.payload should be a monitoring value (usually a numeric).
At configuration time, we must provide some values:
- projectId - The identity of the project that holds the metric definition associated with the value we are recording.
- metricType - The type of the metric we are recording.
Stackdriver constrains the rate at which metrics may be written to a metric. If we attempt to write faster than this, we will get an error:
One or more TimeSeries could not be written: One or more points were written more frequently than the maximum sampling period configured for the metric.
The exact rate isn't clear … it is believed to be no more than one write every 10 seconds.