A Multi-agent Framework for Resource Brokering of Multiple Concurrent Jobs in Grid Environment

In this paper we present the design and implementation of a multiagent framework for executing multiple concurrent jobs in a grid environment. Resource broker plays an imperative role in our framework. The aim of the resource broker is to find, select, reserve and allocate suitable resources to each job submitted by the client. Resource broker always tries to optimize resource utilization and ensure fairness. The multiagent framework assists in executing multiple concurrent jobs in an adaptive execution environment. Services offered by the framework include resource brokering, performance analysis, local tuning and more importantly rescheduling depending on the changing resource requirements. Part of the framework makes use of mobile agents for scheduling and rescheduling purposes. The paper discusses the implementation of the resource brokering and also describes the implementation of a part of the framework which schedules and reschedules the jobs and manages the resources on the basis of some performance monitoring data