基于客户分类的海量数据动态路由存储策略研究  

Dynamic routing storage strategy of mass data based on account classification

作　　者：刘文华 Liu Wenhua(Beijing Wuzi University,Beijing 100000,China)

机构地区：[1]北京物资学院,北京100000

出　　处：《计算机时代》2023年第6期7-10,共4页Computer Era

摘　　要：电商行业的海量数据存储,多采用以客户为切分键进行分库分表的策略,但由于大客户和普通客户数据体量差距巨大,往往会导致数据在各库表中分布严重不均。本文基于分布式数据库存储,设计了区分大客户和普通客户的分库分表策略、能够及时识别客户数据变化的实时数据监控,以及基于客户分类动态选择存储策略的动态路由组件,可有效保障数据的均匀分布。The mass data storage in e-commerce and other industries mostly adopts the strategy of taking the account as the sharding key to divide databases and tables.However,due to the huge gap in the data volume between key account and ordinary account,the distribution of data in each database and table is often seriously uneven.Based on distributed database storage,a customized strategy of divide databases and tables for key account and ordinary account is designed.The real-time data monitoring can identify account data changes in time,and the dynamic routing component can dynamically select storage strategy based on account classification.It can effectively ensure the uniform distribution of data.

关 键 词：客户分类 海量数据 分库分表 动态路由策略 实时数据监控 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]