【作者】 陈珺； 李波； 赵东风；

【Author】 Chen Jun Li Bo Zhao Dongfeng (School of Information Science and Engineering Yunnan University Kunming 650091 China)

【机构】 云南大学信息学院；

【摘要】 在支持网格资源预留的单机调度环境中,没有被预留的时间段对本地作业才是可用的。本文研究如何在单机环境中支持网格作业提前预留的同时,将本地作业合理的分配到不连续的可用时间段,从而提高本地作业执行效率和资源利用率的问题。本文将此调度问题转化为一种新型的装箱问题,提出了松弛时间的灵活调度机制,给出了四种在线调度算法并对这些算法的性能进行了比较研究,结果表明松弛时间机制可以提高本地作业执行效率和提高资源利用率,四种算法中的降序下次适合算法的改善效果最好。更多还原

【Abstract】 Resources Advanced Reservation(AR) make it possible to guarantee the QoS of Grid.However,AR will cause the processing time slices discontinuous and reduce the utilization of local resources.Single machine scheduling is the basis of more complicated parallel machine scheduling.The paper transformed the scheduling problem that allocated the discontinuous time slices for local jobs to a new bin packing model and gave four online scheduling algorithms.The paper proposed a flexible resources reservation mechanism for improving resources utilization of local jobs in the single machine scheduling.Evaluation results show that flexible resources reservation mechanism can improve local resources utilization,and improvement of next fit decreasing algorithm performance is the best.更多还原