计算机科学与探索 ›› 2017, Vol. 11 ›› Issue (9): 1398-1404.DOI: 10.3778/j.issn.1673-9418.1608049

• 数据库技术 • 上一篇    下一篇

海量遥感数据分布式集群化存储技术研究

季  艳1+,鲁克文2,张英慧1   

  1. 1. 北京市遥感信息研究所,北京 100192
    2. 中国电子科技集团公司 第十五研究所,北京 100083
  • 出版日期:2017-09-01 发布日期:2017-09-06

Research on Distributed Clustering Storage Technology for Massive Remote Sensing Data

JI Yan1+, LU Kewen2, ZHANG Yinghui1   

  1. 1. Beijing Remote Sensing Information Institute, Beijing 100192, China
    2. The 15th Research Institute, China Electronics Technology Group Corporation, Beijing 100083, China
  • Online:2017-09-01 Published:2017-09-06

摘要: 针对当前高分辨率遥感数据的高效存储与高速访问迫切需求,采用分布式架构、对象存储和集群技术,结合遥感数据的空间特性,构建了基于数据对象的存储组织模型,设计了全分布式的存储管理架构;形成了逻辑上全球覆盖,物理上分散存储,全球遥感数据存储视图一体化,数据高效共享的分布式集群化遥感大数据存储体系。通过使用此架构,可实现遥感数据资源配置的灵活化,业务区域化特征的定制化与个性化,以及管理系统的智能化。

关键词: 遥感数据, 高性能存储, 分布式集群化, 对象存储

Abstract:

For the urgent needs of the high-resolution earth observation remote sensing data storage and the high-speed access, this paper uses the distributed architecture, object storage and clustering technology, combines the spatial characteristics of remote sensing data to build organizational model based on the data object, designs a storage management architecture with the spatial position as the leading. A remote sensing big data storage system with efficient sharing of distributed cluster can be reformed, that is a global coverage in logic, a dispersed storage in physics and an integration of the storage views of global data. By this architecture, the flexibility of remote sensing data resources configuration, the customization and personalization of business regionalization, and the intelligence of management systems can be achieved.

Key words: remote sensing data, high-performance storage, distributed clustering, object storage