摘要
针对服务网格环境中资源的动态性,提出了一种并行调度算法PGSWA(parallel grid service workflowscheduling),该算法引入了性能预测模型和并行就绪队列来预测下一段时间资源的性能并使得成员服务能够并行执行。实验证明,该算法能较好地缩短工作流的执行时间,提高工作流的执行性能。
Considering the dynamics of resources in the service-oriented grid environment, this paper proposed a novel scheduling algorithm PGSWA( parallel grid service workflow scheduling) to solve such problem. The algorithm introduced the performance prediction model and parallel ready queue into the scheduling to predict the resource performance and run services in parallel. The experiment results show that it can reduce the executing time of workflow, and improve its performance.
出处
《计算机应用研究》
CSCD
北大核心
2008年第11期3285-3287,3295,共4页
Application Research of Computers
基金
国家自然科学基金资助项目(60673041)
国家"863"高技术研究发展计划资助项目(2006AA01Z147)
江苏省高技术研究资助项目(BG2005020)
关键词
网格服务工作流
动态调度
性能预测模型
并行队列
grid service workflow
dynamic scheduling
performance prediction model
parallel queue