摘要
为提高Hadoop云计算平台的性能,该文提出了一种跨层的参数优化模型.首先分析了云计算平台的工作流程,将系统参数与流程对应,并加入基础设施即服务与平台即服务层的参数,找出对Hadoop集群效率作用显著的参数,并把这些参数值作为性能参数,构建成性能参数模型,再用启发式蚁群算法搜寻性能较优的可行参数,并不断修正,找出最佳参数组合,最后整合跨层的参数来提高Hadoop云计算平台的性能.实验表明,该算法可行,性能优良.
In order to improve the performance of the Hadoop cloud computing platform,aparameter optimization model is presented for across layer of IaaS,PaaS and SaaS.Firstly,the work flow of cloud computing platform is analyzed,to make the system parameters correspond to the flow.The parameters of IaaS and PaaS layer are added and parameters which significant impact on Hadoop execution time are found.Performance parameter model is formed based on the above parameters and optimized by Heuristic ant colony algorithm.Finally,cross layer parameters are integrated to improve the performance of Hadoop cloud computing platform.Experiments show that the algorithm is feasible and perform well.
出处
《华中师范大学学报（自然科学版）》
CAS
北大核心
2016年第2期183-189,共7页
Journal of Central China Normal University：Natural Sciences
基金
国家自然科学基金项目(61472135))
广东省科技计划项目(2014B010102007)
