计算机科学与探索 ›› 2012, Vol. 6 ›› Issue (8): 673-683.DOI: 10.3778/j.issn.1673-9418.2012.08.001

• 学术研究 • 上一篇    下一篇

MyBUD自适应分布式存储管理的设计与实现

周宁南1,2,张  孝1,2+,孙新云1,2,琚星星1,2,刘奎呈1,2,杜小勇1,2,王  珊1,2   

  1. 1. 中国人民大学 数据工程与知识工程教育部重点实验室,北京 100872
    2. 中国人民大学 信息学院,北京 100872
  • 出版日期:2012-08-01 发布日期:2012-08-06

Design and Implementation of Adaptive Storage Management System in MyBUD

ZHOU Ningnan1,2, ZHANG Xiao1,2+, SUN Xinyun1,2, JU Xingxing1,2, LIU Kuicheng1,2, DU Xiaoyong1,2, WANG Shan1,2   

  1. 1. Key Laboratory of Data Engineering and Knowledge Engineering, Ministry of Education, Renmin University of China, Beijing 100872, China
    2. School of Information, Renmin University of China, Beijing 100872, China
  • Online:2012-08-01 Published:2012-08-06

摘要: 面对日益增长的非结构化数据管理需求,实现了基于“自由表”数据模型和BUD(bank of unstructured data)参考体系模型的非结构化数据管理平台MyBUD系统。提出了一种能够根据非结构化数据的类型和访问特点自适应地选择分布式存储子系统的方法,同时也对MyBUD进行了TPCC测试和非结构化数据存取实验。结果表明,这种自适应的数据存储方法为MyBUD系统提供了高效的可扩展存储层,为采用数据库方法实现对结构化和非结构化数据统一管理的进一步研究工作奠定了基础。

关键词: 非结构化数据管理, 自适应算法, 分布式存储系统, 面向服务架构

Abstract: Facing with the ever-growing demand for managing unstructured data, this paper proposes a free-table data model and a reference framework BUD (bank of unstructured data), and further implements a prototype platform MyBUD based on the free-table and BUD. It also proposes a novel method to adaptively select a storage sub-system in MyBUD, which manipulates the specific type of unstructured data more efficiently according to its features. In experiment, the paper sets up a TPCC benchmark and a large volume of unstructured video data to evaluate its performance while managing different co-existing categories of data. The experimental results show that the proposed algorithm provides efficient storage service and establishes foundation for further research on unified management of structured and unstructured data in the database approach.

Key words: unstructured data management, adaptive algorithm, distributed storage system, service-oriented architecture