云计算环境下基于并行计算熵的负载均衡算法

发布时间：2018-06-10 17:00

本文选题：云计算 + 虚拟机　；参考：《计算机测量与控制》2014年05期

【摘要】：针对云计算环境下大量并行任务运行所导致的某些节点负载过重,从而引起整个系统负载不均和效率低下的问题,提出了一种基于并行计算熵的资源负载均衡算法;首先,描述了云计算虚拟机部署原理并给出了适合云计算环境和异构集群的并行计算熵的计算方式,然后,定义了在系统并行计算熵低于阈值时迁移的源物理节点、迁移虚拟机和迁移目标物理节点的确定方式;最后,定义了基于并行计算熵的负载均衡算法;采用CloudSim云计算仿真工具对文中方法进行仿真实验,结果表明文中方法较其它方法的平均负载均衡度约低21.8%,具有较低的任务平均响应时间、合理的资源利用率和较小的负载均衡度,具有较大的优越性。
[Abstract]:A resource load balancing algorithm based on parallel computing entropy is proposed to solve the problem that some nodes are overloaded due to the running of a large number of parallel tasks in cloud computing environment, which leads to the uneven load and low efficiency of the whole system. Firstly, a parallel computing entropy based resource load balancing algorithm is proposed. This paper describes the deployment principle of cloud computing virtual machine and gives the computing method of parallel computing entropy suitable for cloud computing environment and heterogeneous cluster. Then, the source physical node is defined when the parallel computing entropy of the system is lower than the threshold. Finally, the load balancing algorithm based on parallel computing entropy is defined, and the method is simulated by CloudSim cloud computing simulation tool. The results show that the average load balancing degree of the proposed method is about 21.8 lower than that of the other methods, and the average task response time is lower, the reasonable resource utilization ratio and the load balancing degree are smaller, and the method has more advantages.
【作者单位】：常州信息职业技术学院;
【分类号】：TP393.01

【参考文献】