The PingDataMetrics Server uses a Monitoring Threshold mechanism that has two components.
Spike monitoring threshold
Used to configure a set of operational performance limits on a specific measurement, where the limit is specified as a percent change from the most recent measurement average value. A Spike Monitoring Threshold has warning and critical limits, and will enter or leave an alerted state when the monitored value exceeds either of the limits. This threshold is useful when the valid range of the measurement is not known in advance. This type of limit is useful in detecting short-term changes in a measurement that fluctuates broadly over time. The limit is applied in both positive and negative directions, so that this type of threshold can detect an upward or downward spike in the value. The following chart shows a spike monitoring threshold, where the red line is the average throughput/second and the green lines are the limits, showing the average window for the throughput.