Spotlight uses metrics, thresholds, and severities to determine the performance statistics of a system. Icons, flows, and labels graphically display this information in the main application window.
A metric is a measurable value resulting from a request made against the target machine. This query runs each time the main application window is refreshed.
The query provides a performance statistic about the system (for example, the number of machines that have at least one server session), and returns a value to Spotlight.
Each metric is divided into thresholds. Some metrics may have only one threshold, while others can have many. A threshold is simply a range of values that can be returned by the metric. A threshold cannot belong to more than one metric.
Each threshold is associated with a severity. The severity determines what action Spotlight is to take when the value returned by a metric falls into the range controlled by a threshold. For example, you may set a severity to display a color, flash the color, emit a sound, or perform an action.
Alarms are attached to severities.
The following diagram shows an example of the relationship between metrics, thresholds, and severities.
In the previous example, the query (% of users connected) is run against the database, and any values returned can fall into one of three ranges (called thresholds). If the query returned the value 20%, the value would fall into the range covered by Threshold 1.
In this example, Threshold 1 is attached to Severity 2. You can set the colors of the severity via the Spotlight Console Options window (for example, Severity 2 might be set to display in green). The component that the metric belongs to then changes to green whenever that threshold is exceeded.