Abstract
|
Real-time processing contains generally chains of complex, heterogeneous, and critical processes. Different monitoring programs are used for detecting anomalies, failures and tracing workflow progress of the running processes. However, the majorities of the traditional monitoring functions of the workflow systems are usually platform dependent, task specific, lacking alerts capabilities, inflexible, and consuming many of the organization resources. This paper introduces the design and implementation of a generic knowledge-based agent model that overcomes these limitations and restrictions. The proposed intelligent agent is using customized rules for workflow monitoring and generating alerts as well as exception reports to the operators. The suggested rule-based agent is fast, autonomous, configurable, reactive, and platform-independent.
|