基于 HDFS管理 MapGIS K9瓦片地图集的研究与实现

摘要

关系型数据库在管理海量空间数据时遇到诸多瓶颈，HDFS（Hadoop Distributed Filesystem）通过将大数据分割为多个小数据块，并将其分别存储在多个数据节点组成的分布式集群中，成为一种新的存储海量空间数据的方法。使用HDFS作为平台，利用数据结构类型MapFile设计一种管理MapGIS K9瓦片地图集的方式，并进行实验。实验表明用HDFS管理海量瓦片地图比传统方式管理更易扩展，更加安全，效率更加高。%Relational databases encounter quite a few bottlenecks while managing the massive spatial data .HDFS becomes a new method to store huge amounts of spatial data by dividing the large data into several small data blocks and separately storing them into several distributed clusters that are composed of data nodes .We design a MapGIS K9 tile map set management means by using HDFS as the platform and utilise the data structure type of MapFile , and make the experiment as well .Experiment shows that to use HDFS to manage massive tile map is more scalable , more secure and has higher efficiency than the traditional ways do .