Research On Building Autonomic Computing Architecture Based On JXTA P2P

Autonomic Computing is a kind of technology which guarantees the service availability of E-commerce infrastructure. Its ultimate goal is to make information systems manage themselves automatically, and maintain their reliability. JXTA is a new platform of network programming and computing, aiming to solve problems existing in modern distributed computing especially P2P computing.This paper expatiates the basic concept, essential, system structure and relations among different components of Autonomic Computing, and based on introduction of traditional P2P, exhibits the specific P2P spark plugged by JXTA from several aspects including the design notion, comparison with other P2P technologies, functional layers, basic members and protocol stack. After analyzing the rationality of combining P2P and Autonomic Computing, and according to the deep understanding of the essence of the two, this paper advances a new angle of view to survey Autonomic Computing, by set up an assumption of treating every component in Autonomic Computing as a peer or a peer group in JXTA world. Contrasting JXTA P2P with Autonomic Computing, this paper also finds that they have quite a lot in common, including system definition, resource identifying, component granularity, uniform access to resources, as well as the system attributes of self-healing, self protecting and environment sensing. Based on the above analysis and the concept of abstracting components in JXTA style, this paper further designs a prototype of "Touchpoint" - one of Autonomic Computing's fundamental components - built on DB2. In this prototype, three peer groups and four peers are created to represent different functional parts of a touchpoint. And they use JXTA services such as pipes and advertisements to carry out "get" and "set" operations successfully, which proves the feasibility of combining JXTA and Autonomic Computing.