通过系统架构的优化实现企业数据仓库的平稳迁移  

Realize the Smooth Migration of Enterprise Data Warehouse Through the Optimization of System Architecture

作　　者：王小玲 

机构地区：[1]中国海洋石油集团有限公司信息技术中心

出　　处：《自动化博览》2022年第9期41-45,共5页Automation Panorama1

摘　　要：本文通过对企业数据仓库体系架构的分析,开展了对模型设计和数据迁移路径的研究,发现可以通过系统架构的优化有效地支持企业数据仓库中数据模型和历史数据迁移的顺利完成。此架构优化的好处也体现在保证了数据分析服务的连续性,业务不会受到企业数据仓库升级期间模型传输和冻结期的影响。另一个关键点就是企业数据仓库底层的分布式大规模并行数据库所提供的多租户云架构。基于此共享架构所搭建的企业数据仓库可以灵活部署数据应用项目,在升级和迁移过程中,利用合理资源部署所需要的开发和测试环境,不仅满足了独立性的要求,也能减轻数据冗余。多租户的系统架构除了满足迁移和升级外,也可以在资源发生瓶颈时进行横向扩展。不仅在系统架构和部署方面应用多租户,通过探索简化数据冗余,更好地履行数据的单一来源的原则方面也可以探索和应用这种系统架构。从而提供用户更好的应用效果、更低的运营成本和最大化的资源投入。Through the analysis of the EDW architecture,the research on model design and data migration path is carried out,and it is found that the optimization of the system architecture can effectively support the smooth completion of the data model and historical data migration in the EDW.The benefits of this architecture optimization are also reflected in the continuity of data analysis services,and the business will not be affected by the model transfer and freezing period during the upgrade of the EDW.Another key point is the multi-tenant cloud architecture provided by the distributed massively parallel database underlying the EDW.The Enterprise Data Warehouse built based on this shared architecture can flexibly deploy data application projects.During the upgrade and migration process,the development and testing environment required by the deployment of reasonable resources can not only meet the requirements of independence but also reduce data redundancy.In addition to migrating and upgrading,the multi-tenant system architecture can also scale horizontally when resource bottlenecks occur.Multi-tenancy can be explored and applied not only in terms of system architecture and deployment but also in terms of the principle of simplifying data redundancy and better fulfilling the single source of data.To provide users with better application effects,more economical operating costs and maximized resource investment.

关 键 词：企业数据仓库 数据拱顶 分层可伸缩架构 租户 多租户数据库容器 大规模并行处理系统 

分 类 号：TP311.52[自动化与计算机技术—计算机软件与理论] F272[自动化与计算机技术—计算机科学与技术]