Design And Some Of The Autonomic Computing System Architecture Based On Agent

As the complexity of the computing systems increase, IT staffs advance a new concept - the Autonomic Computing. Autonomic computing embeds the complexity of the computing systems within them, and realizes the self-managing, that is computing system manages itself, and the managing costs can be reduced.This paper tries to research the autonomic computing from the point of architecture. This paper analyses the architecture of the traditional computing systems, and studies the correlative concepts and mechanisms of the autonomic computing, studies the existing architecture of the autonomic computing, finding that there are some deficiencies such as the insufficiency of the functions and the indeterminacy of the dynamic characteristic. Then we advance an agent-based autonomic computing architecture according to the agent architecture means. In this architecture, design a middle component for the autonomic element to interact, research the structure and the dynamic characteristic of the autonomic element. This architecture has the features: the extensibility, the collaboration and so on. The monitor component is one of the important components of the architecture. It is the basic of the architecture, so at the end of the paper we design and realize it. And we implement a resource monitoring system, which can be customized and the monitoring nodes can be added and deleted at any time, finally we give the experimental monitoring data.