摘要
目前,关系数据库中的分区技术应用相当广泛,但是用分区策略管理海量要素图层数据的存储与索引没有比较系统的技术方法。采用不同管理方式、不同分区粒度、不同索引方式及其组合的分区技术来系统地管理海量空间图层数据,进一步研究了不同的分区粒度及索引方式对查询效率的影响,并通过实验验证了关系数据库中的分区技术对海量要素图层数据的存储与管理具有优化作用。结果表明,在不使用分区键作为查询条件时,分区粒度越大查询效率越高;使用分区键作为查询条件时,本地分区索引查询效率更高等。利用合理的分区方案使得海量要素图层数据存储和管理得以优化,对矢量大数据的存储和管理研究具有重要意义,为更好地应用分区技术来解决实际遇到的存储与检索效率问题提供决策支持。
At present,the partitioning technology in relational database is widely used,but there is no systematic technical method for managing the storage and indexing of massive feature layer data by partitioning strategy.The partitioning techniques of different management methods,different partition granularities,different indexing methods and their combinations are used to systematically manage massive spatial layer data,and the effects of different partition granularity and indexing methods on query efficiency are further studied,and the relationship is verified by experiments.The partitioning technology in the database optimizes the storage and management of massive feature layer data.The results show that the query efficiency is higher when the partition key is not used as the query condition.When the partition key is used as the query condition,the local partition index query is more efficient.Using reasonable partitioning schemes to optimize the storage and management of massive feature layer data is of great significance for the storage and management of vector big data,and provides a decision for better application of partitioning technology to solve the storage and retrieval efficiency problems actually encountered.
作者
叶云霜
林伟华
刘福江
董晓莹
YE Yun-shuang;LIN Wei-hua;LIU Fu-jiang;DONG Xiao-ying(School of Information Engineering,China University of Geosciences,Wuhan 430000,China)
出处
《计算机技术与发展》
2020年第1期167-173,共7页
Computer Technology and Development
基金
国家“十二五”“863”项目(2014AA121401)
湖北省2014年面上自然科学基金项目(2014CFB911)
关键词
海量要素图层
关系数据库
分区
存储
索引
massive feature layer
relational database
partition
storage
index
