Designing and Managing Behavior Models
-
Debugging a Behavior Model - Matching Triggers and Alarm Transitions -
When a trigger is fired, NerveCenter must decide whether that trigger should cause a state transition in an active alarm instance or cause a new alarm instance to be created. What conditions must a trigger and transition meet before one of these actions takes place?
In an active alarm, a transition is pending if its origin state is the alarm instance's current state. A transition is also considered pending if its origin state is Ground. When the second type of transition occurs, a new alarm instance is instantiated.
Triggers have four-part identities. These identities include a name, a subobject, a node, and sometimes a property. Transitions' identities have the same four parts, plus a fifth part, scope. NerveCenter uses matching rules to compare a trigger's identity to the identity of each pending alarm transition. Each pair of names, subobjects, nodes, and properties must pass a comparison test before a transition takes place.
This section describes the identities of triggers and transitions, specifies the matching rules, and provides examples of objects that match and objects that don't match. See the subsections listed below:
The components of a trigger's identity may be supplied by you, the designer, or by NerveCenter, depending on how the trigger is generated. On the other hand, a transition's identity is inherited from an active alarm instance or, if the transition's origin state is Ground, from an alarm definition. The remainder of this section discusses how the components of a trigger or transition's identity are given values.
For triggers fired as a result of a call to the Fire Trigger () function or by a Fire Trigger alarm action, you specify the subobject when you call the function or define the alarm action.
Checking a Poll's Poll Condition | Rules for Matching |
29 July 2003 |