一种云计算系统的瓦片数据组织方法  

A tile data organization method based on Hadoop-pvfs cloud computing system

作　　者：何涛[1] 孙忠芳[1] 葛亮[1] 

机构地区：[1]天津市测绘院,天津300381

出　　处：《测绘科学》2015年第11期164-167,共4页Science of Surveying and Mapping

摘　　要：针对云计算系统中文件数据存储效率较低的问题,文章以Hadoop云计算系统为研究对象,改进了Hadoop-pvfs分布式系统,以提高Hadoop底层文件系统的存储效率,同时使得数据的更新可以在文件的任何位置;并提出一种多层次、位置相关的瓦片数据组织方法。实验结果表明,上述方法相对于Hadoop瓦片服务器系统吞吐效率有较大提高,可以为建设高性能、海量存储的GIS数据库服务系统提供参考。Aiming at the problem that the data storage efficiency is lower,the paper studied Hadoo PCloud Computing system,and improved Hadoop-pvfs distributed system to enhance the storage efficiency of Hadoo Punderlying file system and to update data at any location of the file.Moreover,a multilevel and position related tile data organization method was presented.Experimental result proved the feasibility of the method,which could provide a reference for establishing the GIS service system with high performance and mass storage.

关 键 词：瓦片存储 HADOOP PVFS 云计算 数据组织 

分 类 号：P208[天文地球—地图制图学与地理信息工程]