基于openstack的云计算实训平台  

Cloud computing training platform based on openstack

作　　者：杨慧娟[1] 侯虎 高程 YANG Huijuan;HOU Hu;GAO Cheng(Yulin Vocational And Technical College,Shaanxi Yulin 719000,China;CHA Energy YU LIN Energy Limited Liability Company,Shaanxi Yulin 719000,China)

机构地区：[1]榆林职业技术学院,陕西榆林719000 [2]国能榆林能源有限责任公司,陕西榆林719000

出　　处：《自动化与仪器仪表》2023年第7期161-164,169,共5页Automation & Instrumentation

基　　金：陕西省榆林市科技局2020年产学研项目《榆林市环境监测实验室——基于OpenStack的云计算实训平台关键技术研究》(CXY-2020-017-03)。

摘　　要：对基于OpenStack的云计算实训平台进行了研究,为了解决云计算实训平台资源调度方案设计不合理而出现的资源利用率低、能耗过高等问题,分别对平台整体的物理架构与逻辑架构进行设计,并提出了一种基于HGQP算法的最佳放置方案求解方法,继而获取资源利用率最高、能量消耗最低的虚拟机调度方案,来实现提高云计算实训平台资源利用率和降低能耗的目的。最后分别对HGQP算法与遗传算法、量子粒子优化算法进行实验对比,并对基于OpenStack的云计算实训平台进行了功能测试。结果表明:HGQP算法适用于虚拟机调度问题,当虚拟机增加到300时,最优调度方案获取时间仅需16 s,CPU利用率为0.45,能量消耗为0.81,在三种算法中计算速度最快、耗时最短、稳定性更高,且当虚拟机请求增加越多,HGQP算法的优势更强。同时,基于OpenStack的云计算实训平台的各个功能模块能够反应迅速并正常运行。The cloud computing training platform is studied based on OpenStack,in order to solve the unreasonable design of low resource utilization and high energy consumption,respectively for the overall physical architecture and logic architecture design,and a best placement is proposed based on HGQP algorithm solution method,and then obtain the highest resource utilization,the lowest energy consumption of virtual machine scheduling scheme,to achieve the purpose of improving cloud computing training platform resource utilization and reduce energy consumption.Finally,the HGQP algorithm and the OpenStack-based cloud training platform were experimentally verified and tested.The results show that the HGQP algorithm can obtain the optimal scheduling scheme of virtual machines that meet the requirements,and the cloud training platform of OpenStack can run quickly and stably,and its functional modules can reflect the rapid normal operation.

关 键 词：OPENSTACK 云计算实训平台 HGQP算法 虚拟机 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]