计算机与现代化

• 数据库与数据挖掘 • 上一篇    下一篇

基于云存储的冗余分片自适应分发策略

  

  1. 河海大学计算机与信息学院,江苏南京211100
  • 收稿日期:2015-08-11 出版日期:2016-01-22 发布日期:2016-01-26
  • 作者简介:查显月(1989-),男,安徽安庆人,河海大学计算机与信息学院硕士研究生,研究方向:时空数据管理; 冯钧(1969-),女,教授,博士生导师,研究方向:时空数据管理,智能数据处理与数据挖掘,水利信息化; 冯读庆(1989-),男, 硕士研究生,研究方向:云存储与云计算; 唐志贤(1983-),男,博士研究生,研究方向:时空数据管理。
  • 基金资助:
    国家自然科学基金资助项目(61370091, 61170200)

An Adaptive Distribution Strategy of Redundant Splits on Cloud Storage

  1. College of Computer and Information, Hohai University, Nanjing 211100, China
  • Received:2015-08-11 Online:2016-01-22 Published:2016-01-26

摘要: 水利信息化的发展促进水利大数据的快速增长,在实际生产中发现,水利数据的整合与共享能充分发挥水利大数据的综合效益,但由于各数据中心存储能力不均衡等问题,很容易产生数据交换热点。本文在混合存储模型的研究基础上,根据数据中心之间交换需求量以及数据中心存储能力,设计基于水利数据领域空间特性的冗余分片自适应分发策略,并利用该策略进行数据分片的分发决策。实验表明,该策略能够基本保证各数据中心存储负载趋于均衡。

关键词: 云存储, 数据交换, 负载均衡, 冗余分片

Abstract: The water conservancy informatization evolution promotes the rapid growth of water conservancy big data. And the actual production has proved the integration and sharing of the data can achieve the maximum benefit of them. But the existing storage platforms have a few deficiencies, such as unbalanced of storage capacity, and that can make these platforms be a hotspot. Based on the research of hybrid storage model, according to the exchange demand between the data centers and the storage capacity, we designed an adaptive distribution strategy based on spatial property of water conservancy data domain in which conducting the distribution decision. The experiment shows that this strategy can guarantee the basic load-balancing between data centers.

Key words: cloud storage, data exchange, load-balancing, redundant splits

中图分类号: