摘要
在商业网格和云计算环境中,作业有到达时间、计算量、预算、截止期等参数,其中,预算是时间的函数。准确区分作业的重要性和紧迫性是作业调度系统的一个关键问题。综合利用这四个参数来定义作业的优先级,并提出基于价值密度和相对截止期的网格作业调度算法。分别对弱实时和强实时网格作业的调度进行仿真。仿真结果显示,所提出的调度算法的性能在两种情况下都优于所有对比算法的性能,且在强实时作业情况下优势更明显。
Jobs have arrival time,workload,budget and deadline parameters in economy-based grid computing and cloud computing environment.And the budget parameter changes over time.It is vital to differentiate jobs＇ importance and urgency exactly for job scheduling system.These four parameters are all taken into account in defining a new priority of jobs here. Based on this priority, a value density and relative deadline jointly based scheduling algorithm is proposed.Results of simulation show that the proposed scheduling algorithm outperforms the referenced algorithms in scheduling of soft and hard realtime grid jobs,especially in the latter.
出处
《计算机工程与应用》
CSCD
北大核心
2010年第24期1-3,21,共4页
Computer Engineering and Applications
基金
国家重点基础研究发展计划(973)(No.2007CB307103)
贵州省重大科技专项计划(黔科合重大专项)(No.【2007】6017)~~
