摘要
由于资源具有广域分布、异构、动态等特性,计算网格环境下资源的管理和调度是一个非常复杂且具有挑战性的问题.提出了计算网格环境下一组相互独立的计算任务(meta-task)的资源映射策略.该策略采用重复映射方法,以更好地适应网格计算环境下的动态性和自治性.算法考虑到任务的输入数据位置对映射效果的影响;通过定义效益函数,该策略在追求较小的任务完成时间的同时兼顾任务的服务质量(QoS)需求.模拟实验结果显示,该映射策略更符合计算网格的复杂环境,能够更好地满足不同用户的实际需要.
The management of resources and scheduling computations in a grid environment is a complex undertaking, mainly due to resource抯 geographic distribution, heterogeneity, distributed ownership with different policies and priorities, varying loads, reliability, and availability conditions. A unified resource mapping strategy in computational grid environments is presented, which considers the input data repositories and QoS of tasks to mapping a set of independent tasks (meta-task) to resources. The repetitive mapping algorithm is more suitable for the dynamic adaptability and domain autonomy in the grid. The benefit function heuristic adopted in the algorithm can assure the QoS of tasks more effectively.
出处
《软件学报》
EI
CSCD
北大核心
2002年第7期1303-1308,共6页
Journal of Software
基金
国家重点基础研究发展规划973资助项目(G1998030403)
