This overview describes the key concepts of the Active Correlation
Technology rule language.
A rule pattern is the representation of an event correlation situation
(such as a threshold condition or duplicate event detection). The Active Correlation
Technology rule language includes seven rule patterns that have been proven
to represent most of the event correlation situations that IBM® customers
need to address. Six of the seven rule patterns define stateful rules, and
one of the patterns defines a stateless rule.
Stateful rules correlate multiple events that occur during a specific time
period and generate a response to those events. Stateless rules process only
a single event that meets a certain condition and generate a response to that
event.
- stateful rule
- A rule that retains state information, which is information about the
characteristics of a rule instance, for the purpose of acting on a collection
of events over a period of time. Rules that are defined by any of the following
rule patterns are stateful rules: collection, computation, duplicate, sequence,
threshold, or timer.
- stateless rule
- A rule that does not retain state information and therefore can act only
on one event at a time. A rule that is defined by the filter pattern is a
stateless rule.