Clusters, grids, and peer-to-peer (P2P) networks have emerged as popularparadigms for next generation parallel and distributed computing. Themanagement of resources and scheduling of applications in such large-scaledistributed systems is a complex undertaking. In order to prove theeffectiveness of resource brokers and associated scheduling algorithms, theirperformance needs to be evaluated under different scenarios such as varyingnumber of resources and users with different requirements. In a gridenvironment, it is hard and even impossible to perform scheduler performanceevaluation in a repeatable and controllable manner as resources and users aredistributed across multiple organizations with their own policies. To overcomethis limitation, we have developed a Java-based discrete-event grid simulationtoolkit called GridSim. The toolkit supports modeling and simulation ofheterogeneous grid resources (both time- and space-shared), users andapplication models. It provides primitives for creation of application tasks,mapping of tasks to resources, and their management. To demonstrate suitabilityof the GridSim toolkit, we have simulated a Nimrod-G like grid resource brokerand evaluated the performance of deadline and budget constrained cost- andtime-minimization scheduling algorithms.
展开▼
