融合架构的分布式数据库技术在气象大数据平台上的应用实践  

Application Practice of Distributed Database Technology with Fusion Architecture on Meteorological Big Data Platform

作　　者：汪华[1] 李波[1] 王彪 廖婷婷 WANG Hua;LI Bo;WANG Biao;LIAO Tingting(Guizhou Meteorological Information Center,Guiyang 550002,China)

机构地区：[1]贵州省气象信息中心,贵州贵阳550002

出　　处：《中低纬山地气象》2020年第5期93-96,共4页Mid-low Latitude Mountain Meteorology

基　　金：黔科合支撑[2019]2386号:基于大数据的气象防灾减灾服务平台构建与应用。

摘　　要：随着信息技术的发展和应用,气象发展已融入到政治、经济、民生等各领域的发展中,气象数据呈现数据体量大、实时性要求高、数据种类丰富的大数据特征,数据的应用效果随时间呈断崖式下降的趋势非常明显,是区别其他行业数据应用的明显差异点。传统关系型数据库暴露出长序列数据访问效率低、软硬件架构成本高、性能扩展能力弱等缺陷,已无法满足业务发展的需要,该文通过融合架构的分布式数据库在气象业务中的应用效果分析,从技术原理和实践两个角度,为气象大数据的应用提供技术借鉴。With the development and application of information technology,meteorological development has been integrated into the development of politics,economy,people's livelihood and other fields.Meteorological data shows the characteristics of big data with large volume,high real-time requirements,and rich data types.The application effect of data shows a cliff like downward trend over time,which is a significant difference between data applications in other industries.The traditional relational database exposes the defects of low access efficiency of long sequence data,high cost of software and hardware architecture,weak capability of performance expansion and so on,which can't meet the needs of business development.This paper analyzes the application effect of the distributed database based on the fusion architecture of meteorological business,and provides technical direction reference to the application of meteorological big data from two aspects of technical principle and practice.

关 键 词：分布式数据库 HADOOP 融合架构 长序列 

分 类 号：TP311.133.1[自动化与计算机技术—计算机软件与理论]