基于服务器集群的大数据实训平台建设

Construction of Big Data Teaching Platform Based on Server Cluster

数据科学与大数据技术专业是“新工科”背景下的新兴专业,其理论教学体系已初步形成,但在实践教学尤其是实验平台建设方面存在不足。为此,提出基于Hadoop集群和虚拟化技术的大数据实训平台建设方案,并开展递进式项目化教学实践。大数据实训平台建设思路为“2+2+2”方案,即构建Hadoop和云计算两个功能集群,部署教学管理和资源虚拟管理两组系统,服务大数据日常教学和创新实践两类应用。梳理大数据专业实践教学体系的递进关系,在大数据实训平台上,以推荐系统为例进行了实验设计和项目教学实践,取得良好教学效果。

Data Science and Big Data Technology is an emerging specialty under the background of New Engineering.Its theoretical teaching system has been preliminarily formed,but there are deficiencies in practical teaching,especially in the construction of the experimental platform.Therefore,this paper puts forward the construction scheme of a big data teaching platform based on Hadoop cluster and virtualization technology,and carries out progressive projected-based practical teaching.The construction scheme is abbreviated as“2+2+2”,namely two functional clusters of Hadoop and cloud computing,two groups of teaching management and resource management systems,and two kinds of applications both daily teaching and innovative practice.Analyzing the progressive relationship of the big data professional practice teaching system,the project teaching is implemented on this platform with the recommendation system as an example.Students’experimental experience and overall teaching effect have been significantly improved.