Research And Application On Policy-based Distributed Resource Monitoring Model

Along with the rapid development of network technologies, heterogeneous resources from independent locations can be integrated to a loose-coupling computing environment, that is, Distributed Computing Environment (DCE). Due to the loose-coupling structure of computing nodes in DCE, resource managements must be emphasized for the uniform view of application services. And resource monitoring is the keystone of distributed resource managements. Data from monitoring can be used to assist job scheduling, load balance, reporting failures of software or hardware and widely controlling the system usage.We present a policy-based resource monitoring model in this paper and use the model in our multi-node intelligent network application environment which is based on virtual personal computing and integrates network applications and various terminal devices to provide clients a self-adaptation, supporting-user-mobility information service platform. We gave prominence to high performance, extensibility, flexibility, reliability and manageability of the model design. The model has several characters: peer-to-peer-monitoring, local-storage and two-dimension-statistics which includes server-based-statistics and user-based-statistics. Moreover, introducing policy control makes a novel perspective for automata and intelligence of the monitoring system.