APP下载

超融合架构的大数据虚拟仿真实验平台建设研究

2020-09-02罗剑

软件导刊 2020年8期

罗剑

摘 要:为了更好地开展大数据实验教学与研究,提高大数据实验课程质量,克服传统大数据实验建设弊端,基于KVM技术与超融合架构设计一种新的大数据虚拟仿真实验平台建设方案。该方案将计算和存储融合在X86单一物理节点中,分布式存储成本低,提高了IO性能,具有开放兼容、易于扩展,方便维护、高可用等特点。该大数据虚拟仿真实验平台能够简化大数据实验平台环境搭建,提高教学实验效率。基于该平台可以定制各种实验环境,通过弹性分配计算资源,保障实验教学环境稳定运行。

关键词:大数据实验;KVM技术;超融合架构;实验平台;虚拟化技术

DOI:10. 11907/rjdk. 201606 开放科学(资源服务)标识码(OSID):

中图分类号:TP319文献标识码:A 文章编号:1672-7800(2020)008-0151-05

Abstract: In order to better carry out the teaching and scientific research of big data experiment, improve the quality of big data experiment course, improve the disadvantages of the traditional big data experiment construction, based on KVM technology and Hyper-converged infrastructure, a new big data virtual simulation experimental platform construction scheme is designed and it integrates computing and storage into the X86 single physical node, distributed storage costs are low. The scheme improves IO performance, has open compatibility, and it is easy to expand, maintaine and of high availability and so on. By using this big data virtual simulation experimental platform, it can simplify the construction of big data experiment platform environment and improve the efficiency of teaching experiment, and under this platform, various experimental environments can be customized, and the scientific resources can be allocated flexibly to ensure the stable operation of the teaching environment.

Key Words: big data experiments; KVM technology; HCI; experimental platforms; virtualization technologies

0 引言

伴随着新一轮科技革命与产业变革,人才培养模式也实现了转型升级。2018年,教育部首批认定了612个新工科研究与实践项目,积极探索“新工科”建设新理念,组建了包括大数据在内的项目群,深入推进新工科建设[1]。由于大数据技术具有多学科交融性[2-3],很多专业都开设了大数据相关课程,或对已有专业课程体系进行升级[4],各高校在积极推进大数据专业建设的同时,也逐步开展大数据实验室建设,以满足大数据实践型、复合型人才需求[5]。

大数据实验教学与研究对计算机硬件、软件、数据等都有很高要求,传统的单一伪分布式、多机集群模式等实验室已无法满足大数据新型技术应用要求,有高校通过使用Docker容器技术减少实验成本,提高教学实验便利性[6-7]。……

登录APP查看全文