电信科学 ›› 2018, Vol. 34 ›› Issue (1): 148-157.doi: 10.11959/j.issn.1000-0801.2018043

• 运营商大数据专栏 • 上一篇    下一篇

基于PaaS技术的大数据云化平台实践

徐海勇,黄岩   

  1. 中国移动通信集团有限公司,北京100032
  • 修回日期:2017-12-01 出版日期:2018-01-01 发布日期:2018-02-05
  • 作者简介:徐海勇(1970-),男,中国移动通信集团有限公司业务支撑系统部总经理、高级工程师,主要研究方向为移动通信、互联网、大数据。|黄岩(1976-),男,中国移动通信集团有限公司业务支撑系统部信息管理处经理、高级工程师,主要研究方向为大数据、能力开放。

Applications of a big data platform based on PaaS

Haiyong XU,Yan HUANG   

  1. China Mobile Communications Groups Co.,Ltd.,Beijing 100032,China
  • Revised:2017-12-01 Online:2018-01-01 Published:2018-02-05

摘要:

大数据平台作为数据存储、处理和服务的基础平台,需支撑内部和外部多种大数据应用的开发及运行。提出一种把PaaS(platform as a service,平台即服务)技术应用到大数据平台的方案,详细描述了大数据PaaS云化平台的架构体系以及分层调度、多租户管理、Hadoop任务全局调度等关键技术方案。实践证明,该平台可以显著提高大数据应用的开发效率,实现应用快速部署上线;同时,其良好的弹性伸缩能力可以有效地提高大数据应用系统资源的使用效率。

关键词: 大数据, PaaS, Hadoop, YARN调度器, 云服务

Abstract:

As a platform of data storage,processing,and service,big data platform needs to support internal and external various big data applications.A solution applied PaaS technology to the big data platform was proposed.Details of the key technical solutions such as the architecture of PaaS cloud computing platform,hierarchical scheduling scheme,multi-tenant management and global scheduling of Hadoop tasks were described.It was proved that the platform could remarkably improve big data application development efficiency and help application deployment fast.In the meantime,good elastic scaling capability can effectively improve big data application system resource utilization.

Key words: big data, PaaS, Hadoop, YARN scheduler, cloud service

中图分类号: 

No Suggested Reading articles found!