网格环境下基于QoS的资源映射策略

QoS-based Resource Mapping Policy in Grid Environment

资源映射是网格计算研究领域中的一个研究热点。现有的研究工作通常考虑任务的完成时间,对于满足用户QoS需求前提下提高资源映射性能方面还缺乏深入的研究。该文提出一种网格环境下相互独立任务的基于QoS的资源映射策略——QSPCR_Sufferage。综合考虑了任务的截止期属性和多QoS参数,选择最优资源。实验结果表明,该策略可以在满足用户多QoS需求下,有效地提高任务请求成功率和资源利用率。

Resource mapping is an emergency problem in grid computing area, current research focuses on how to minimize the makespan and improve the throughput, but the research on how to meet the users＇ QoS need and improve the mapping performance is not very abundant. For that, a QoS-based resource mapping policy（QSPCR_Sufferage） is proposed. The best resource is chosen according to the task deadline attribute and multiple QoS parameters. The result of experiment indicates that this policy can effectively increase the task request success rate and improve the grid resource utilization.