基于OpenStack的资源调度算法研究与改进  

Research and Improvement of Resource Scheduling Algorithm Based on OpenStack

作　　者：郑建忠 郑建荣 王慧 ZHENG Jianzhong;ZHENG Jianrong;WANG Hui(Beijing China Power Information Technology Co.,Ltd.,Beijing 100192,China;Educational Administration Branch,Pingluo Vocational Education Center ofNingxia,Pingluo 753400,China)

机构地区：[1]北京中电普华信息技术有限公司,北京100192 [2]宁夏平罗县职业教育中心教务科,宁夏平罗753400

出　　处：《电力信息与通信技术》2018年第12期46-50,共5页Electric Power Information and Communication Technology

基　　金：国家电网公司总部科技项目资助"云计算在大并发;复杂计算特征下的电网业务场景应用"

摘　　要：为了提升资源利用率、合理分配资源、加快资源调度,文章针对基于OpenStack资源调度算法在动态调整方面的不足,对综合负载均衡调度算法补充任务连接数影响因子,并增加动态权值计算方法,提出改进的综合负载均衡调度算法,解决了算法权值动态调整受限问题,从而实现了高效动态资源调度和资源平衡。通过仿真测试工具CloudSim实验,从资源利用率和负载性能2方面对改进前后的算法结果进行对比,得出改进的综合负载均衡调度算法能够更高效地改善资源调度能力和负载均衡能力。In order to improve resource utilization,allocate resources reasonably and speed up resource scheduling,this paper aims at the shortage of dynamic adjustment of resource scheduling algorithm based on OpenStack,and complements the influence factor of task connection number to comprehensive load balancing scheduling algorithm.The dynamic weight calculation method is added and an improved comprehensive load balancing scheduling algorithm is proposed to solve the constrained problem of dynamic adjustment of the weight value of the algorithm so as to achieve efficient dynamic resource scheduling and resource balance.The results of the improved algorithm are compared from the two aspects of resource utilization and load performance through the simulation test tool CloudSim experiment.It is concluded that the improved comprehensive load balancing algorithm can improve resource scheduling ability and load balancing ability more efficiently.

关 键 词：OPENSTACK 资源调度 动态调度 负载均衡 权值 综合负载均衡调度算法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]