基于负载特征聚类的节能资源调度算法  

Energy-saving resource scheduling algorithm based on workload characteristic clustering

作　　者：夏庆新[1] 兰雨晴[1] 唐甜[1] 肖利民[1] 

机构地区：[1]北京航空航天大学计算机学院,北京100191

出　　处：《北京航空航天大学学报》2015年第4期680-685,共6页Journal of Beijing University of Aeronautics and Astronautics

基　　金：国家自然科学基金重点资助项目(61232009)

摘　　要：基础设施即服务(Iaa S,Infrastructure as a Service)平台提供商为用户提供高性能服务的同时,必须考虑如何在不违反服务级别协议(SLA,Service Level Agreement)的前提下,节约云平台的能耗成本.采用基于负载特征聚类的方法,提出一种Iaa S云平台上保证SLA的资源调度算法,最终实现降低SLA违反率和节约能耗的目标.具体采用能耗相关的负载特征提取和改进K-means聚类分析的研究方法,进行资源调度算法研究,对物理资源进行有效分配,以保证Iaa S平台节约能耗的要求.实验验证方面,通过扩展Cloud Sim模拟实验平台,对本研究算法与改进BFD(Best Fit Decreasing)算法进行比较,得出本研究算法在SLA违反率和节能方面更优.When infrastructure as a service （IaaS） providers offer high performance services for users, they must think about how to reduce the energy cost of the cloud platform without violating the service level agreement （SLA）. A resource scheduling algorithm to ensure SLA was proposed based on clustering analysis of the load characteristic. Ultimately, the targets of reducing SLA violation rate and saving energy were realized. The resource scheduling algorithm was analyzed based on improved K-means clustering analysis and extraction of workload characteristic according to energy consumption. Physical resources were effectively allocated to ensure the requirement of energy saving of IaaS platform. Based on the extension of the CloudSim simulation platform, the algorithm proposed was compared with the optimized best fit decreasing （BFD） to show lower SLA violation rate and energy consumption.

关 键 词：基础设施即服务 负载 聚类 节能 服务级别协议 

分 类 号：V221.3[航空宇航科学与技术—飞行器设计] TB553[理学—物理]