An Efficient Decentralized Grid Service Advertisement Approach Using Multi-Agent System

Grid resource management mainly schedules applications over the available resources. Such scheduling over a large scale distributed environment raises performance issues. Most of the current implementations paid a little consideration of performance issues that are raised at service discovery and advertisement processes. In this work an agent based resource management system, ARMS, is enhanced with an adaptable decentralized service advertisement strategy to reduce the cost of the advertisement process within ARMS system. A Performance Monitoring and Advisory for ARMS component is implemented to compare the developed strategy with those previously defined in literature. Simulated test results recorded a noticeably better performance for the new strategy since it is more adaptable to continuous changes in both workload and structure of ARMS system.