Grid Resource Management Based on Functional Dependency

Abstract

In this paper, we propose a resource management system in Grid computing in order to specify system Quality of Service (QoS) requirements for dynamic and complex emerging applications. Our approach is based on the functional dependency among application components to specify the probability of system QoS requirements for the emerging application. Experimental results show that our application scheduling based on functional dependencies can achieve scheduling and managing emerging applications to satisfy a client’s quality of service in Grid computing. The results also show significant improvement of performance comparing to cluster distribution and random distribution scheduling approaches.