云计算网络下虚拟机负载均衡方法仿真  

Simulation of virtual machine load balancing method in cloud computing network

作　　者：舒晓苓 吴雪琴 SHU Xiao-ling;WU Xue-qin(Chengdu College of University of Electronic Science and Technology of China,Sichuan Chengdu 611731,China)

机构地区：[1]电子科技大学成都学院,四川成都611731

出　　处：《计算机仿真》2022年第3期358-361,412,共5页Computer Simulation

摘　　要：由于不能合理协调网络运行资源,导致通信拥挤、安全性能下降,为此,提出基于逻辑分区的云虚拟机负载均衡优化算法。采用逻辑分区方式划分各分区虚拟机负载情况与业务形式,并以平均任务完成时长当作优化目标,搭建资源调度数学模型,根据虚拟机配置数组情况把多维资源协调问题变换成单维协调问题;利用遗传算法对单维虚拟机资源编码与解码,生成新的资源,运用基因突变原理自适应调节资源,完成负载均衡优化。实验结果表明,所提算法负载均衡效果较好,资源调度时间最短,并具有良好的收敛性。Communication congestion and security performance degradation were caused by unreasonable network operation resources.In this regard,we report a cloud virtual machine load balancing optimization algorithm based on a logical partition.According to the logical partition mode,the load and service form of each partition virtual machine were divided.The average task completion time was taken as the optimization goal.The mathematical model of resource scheduling was founded.According to the virtual machine configuration array,the multidimensional resource coordination was transformed into variable unidimensional coordination.The genetic algorithm was introduced to encode and decode one-dimensional virtual machine resources to obtain new resources.Based on the principle of gene mutation,resources were adjusted adaptively to complete the optimization of load balancing.The experimental results show that the algorithm has an excellent load balancing effect,reduced resource scheduling time and significant convergence.

关 键 词：逻辑分区 云虚拟机 云计算 负载均衡 虚拟机配置 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]