计算机科学与探索 ›› 2016, Vol. 10 ›› Issue (10): 1351-1364.DOI: 10.3778/j.issn.1673-9418.1510042

• 综述·探索 • 上一篇    下一篇

面向数据的软件体系结构初步探讨

苗  放1,2+   

  1. 1. 成都理工大学 地球物理学院,成都 610059
    2. 地质灾害防治与地质环境保护国家重点实验室,成都 610059
  • 出版日期:2016-10-01 发布日期:2016-09-29

Discussion of Data-Oriented Software Architecture

MIAO Fang1,2+   

  1. 1. College of Geophysics, Chengdu University of Technology, Chengdu 610059, China
    2. State Key Laboratory of Geohazard Prevention and Geoenvironment Protection, Chengdu 610059, China
  • Online:2016-10-01 Published:2016-09-29

摘要: 受软件体系结构的限制,信息技术领域长期存在的问题在大数据时代愈发突出:系统间的信息难以共享;已建系统功能扩展困难;海量、异构、多源、动态、实时变化和爆发式增长的大数据难管理,难分析,难挖掘;面向业务的软件工程开发过程复杂,维护困难,生命周期短;在互联网开放环境下的信息安全、数据安全问题面临挑战;数据所有者利益得不到保障等。面向数据的软件体系结构(data-oriented software architecture,DOA),采用“面向数据和以数据为核心”的思想,通过数据注册中心(data register center,DRC)、数据权限中心(data authority center,DAC)和数据异常中心(data exception control center,DEC)统一定义数据、管理数据和提供数据服务;通过数据应用单元(data application units,DAUs)对各种应用进行管理和服务,建立了一种数据大平台与碎片化应用的数据生态系统,为构建大数据时代从数据保护到授权应用整套机制的软件体系结构,进行了有益的探索。

关键词: 大数据, 软件体系结构, 面向数据的体系结构, 数据生态系统, 数据注册中心, 数据权限中心

Abstract: The problems existing in the field of information technology for a long time are more and more serious due to the limitation of software architecture. Information between different systems is burdensome to share, and it is difficult to expand the built-up systems. Besides, data in massiveness, heterogeneity, various sources, real-time, dynamic and in explosive growth, are unmanageable, difficult to analyzing and mining. There are several problems in business-oriented software engineering, notably, it is hard to manage the process of developing and maintaining. The software life-cycle is not usually sustained. What is more, the problem of information security and data security is facing a challenge in network-opening environment. And more, the benefits of data owner will not usually be protected. This paper brings up a new architecture called data-oriented software architecture (DOA), which is designed by the ideology of “orienting data” and “data as kernel”. The DOA systems define, manage data and provide services through data register center (DRC), data authority center (DAC) and data exception control center (DEC). In addition, the data application units (DAUs) are designed to realize the management of various kinds of applications. The destination is to set up a data ecosystem for big data platforms and fragmented applications. This paper explores the architecture which considering the mechanism of data protection and authorization in the age of big data.

Key words: big data, software architecture, data-oriented architecture, data ecosystem, data register center, data authority center