面向大规模云资源调度的可扩展分布式调度方法

发布时间：2018-01-20 14:06

本文关键词： 云计算资源分配能耗优化分而治之 CSP　出处：《计算机工程与科学》2015年11期 　论文类型：期刊论文

【摘要】：云数据中心异构物理服务器的能耗优化资源分配问题是NP难的组合优化问题,当资源分配问题规模较大时,求解的空间比较大,很难在合理时间内求得最优解。基于分而治之的思想,从调度模式方面提出可扩展分布式调度方法,即当云数据中心待调度的物理服务器的数量比较大时,将待调度的服务器划分为若干个服务器集群,然后在每个服务器集群建立能耗优化的资源分配模型,并利用约束编程框架Choco求解模型,获得能耗最优的资源分配方式。将提出的基于可扩展分布式调度方法的能耗优化云资源调度算法与非扩展调度算法进行实验比较,实验结果表明,提出的基于可扩展分布式调度方法的能耗优化云资源调度算法在大规模云资源分配上有明显的性能优势。
[Abstract]:The problem of energy consumption optimization for heterogeneous physical servers in cloud data centers is a NP-hard combinatorial optimization problem. When the scale of resource allocation problem is large, the space for solving the problem is relatively large. It is difficult to find the optimal solution in a reasonable time. Based on the idea of divide-and-conquer, an extensible distributed scheduling method is proposed from the aspect of scheduling mode, that is, when the number of physical servers to be scheduled in cloud data centers is large. The servers to be scheduled are divided into a number of server clusters, and then a resource allocation model of energy consumption optimization is established in each server cluster, and the model is solved by using the constraint programming framework (Choco). The energy consumption optimal resource allocation method is obtained. The proposed energy consumption optimization cloud resource scheduling algorithm based on scalable distributed scheduling method is compared with the non-extended scheduling algorithm. The experimental results show that. The proposed energy consumption optimization algorithm based on scalable distributed scheduling method has obvious performance advantages in large-scale cloud resource allocation.
【作者单位】：华南理工大学计算机科学与工程学院;
【基金】：国家自然科学基金资助项目(61402183) 广东省科技计划资助项目(2014B010117001,2014A010103022,2014A010103008,2013B090200021,2013B010401005) 中央高校基本科研业务费专项资金资助项目(xzjsD2153930) 广州开发区萝岗区科技领军人才项目(2014P176)
【分类号】：TP393.01
【正文快照】： 通信地址:510640广东省广州市番禺大学城华南理工大学计算机科学与工程学院Address:School of Computer ScienceEngineering,South China University of Technology,Guangzhou 510640,Guangdong,P.R.China1引言云计算技术[1,2]是分布式计算、并行计算、效用技术、网络存储等

【参考文献】