A web-oriented agent-based system for monitoring real-time data processing

Real-time data processing systems generally contain a series of heterogeneous, complex and critical processes. These systems use monitoring tools to increase their productivity and efficiency by detecting failures and tracing workflow progress for the different processes. However, most of the current monitoring tools are platform dependent, process/task specific, expensive, difficult to maintain, and consume many of the organization resources. This paper introduces a prototype generic agent-based and task-unspecific system that overcomes these limitations and restrictions. The proposed platform-independent system is structured into three different modules. These modules collect, organize, infer, and visualize the status of the different objects as well as their necessary attributes/properties from the different stages of the monitored system. The proposed agent-based monitoring system is fast, dynamic, easy to configure, and more convenient to implement. The monitoring agent is reactive, autonomous and communicative. Moreover, it can be used either from any internet browser or in a stand alone mode.