摘要
针对云计算数据中心资源分配算法的资源利用率较低的问题,提出了一种基于改进遗传算法的云计算数据中心资源分配算法。首先,遍历每个服务器与虚拟机的需求,使用启发式贪婪算法,按照遗传算法搜索的最优虚拟机顺序将虚拟机分配至物理服务器;然后,将染色体对应的虚拟机顺序转化为装箱问题的装箱解,根据适应度值搜索资源池的最优顺序;最终,最小化云计算物理服务器的数量,减少了物理服务器的资源浪费量。基于不同虚拟机规模进行了仿真实验,结果显示:本算法对于多维装箱问题具有较好的性能,在云计算资源分配方面也获得了较好的资源利用率。
Concerning the problem that the current resource allocation algorithms of cloud computing data center perform low resource utilization ratio,a greedy resource allocation algorithm of cloud computing based on improved genetic algorithm is proposed. Firstly,each server and the requirement of each virtual machine are scanned,heuristic greedy algorithm is used to assign the virtual machines to each physic server according to the optimal virtual machine order optimized by genetic algorithm.Then,the virtual machine orders corresponding to the chromosomes are transformed to the packing solution of packing problem,the optimal order of resource pool is searched according to the fitness values. Lastly,the amount of physic servers of clouding computing is minimized and the resource wastage of physic servers are reduced. The simulation experimental results based on different virtual machine scales show that the proposed algorithm outperforms the other algorithms to multi dimensional packing problems,and realizes a better resource usage ratio than the other algorithms to cloud computing resource allocation application.
作者
蔡柳萍
俞龙
CAI Liuping;YU Long(School of Computer Science & Engineering, Tianhe College of Guangdong Polytechnic Normal University, Guangzhou 510540,China;College of Electronic Engineering, South China Agricultural University, Guangzhou 510642, China)
出处
《重庆理工大学学报(自然科学)》
CAS
北大核心
2019年第4期134-140,共7页
Journal of Chongqing University of Technology:Natural Science
基金
广东省工业机器人集成与应用工程技术研究中心专题科研项目"人机交互技术开发与应用"(201812GCZX003)
广东技术师范学院天河学院计算机科学与技术重点学科建设项目(Xjt201702)
广东高校特色创新类项目"物联网智能设备云数据中心的应用与研究"(2015KTSCX184)
关键词
数据中心
资源分配
虚拟机部署
遗传算法
多维装箱问题
贪婪算法
cloud computing
resource allocation
virtual machine deployment
genetic algorithm
multi dimensional packing problem
greedy algorithm