数据库系统应用分片中间件

随着国家电网信息化系统一级部署的陆续实施, 业务数据库的表越来越多, 表中的数据量越来越大. 海量数据的存储与访问对数据库造成了相当大的负载, 数据处理能力和访问能力均遭遇瓶颈. 本文针对国家电网一级部署数据量快速增长的特点, 提出了数据分片中间件UAPDS(UAP DB Sharding). UAPDS以集群、分区、分片三层拆分模式, 通过读写分离、单库分表、多库分表等多种解决方案将数据按照一定的规则切分到不同的库或表中, 有效的缓解了对数据库的压力, 提高了数据访问效率, 在实际应用中取得了较好的效果.

Middleware of Database System Fragmentation

As the successively implementation of State Grid''s information system first-degree deployment, the tables in database become much more than before, the datas in table become very large. Massive data storage and access caused considerable load on the database, data processing capacity and access encounter bottlenecks. The amount of data is growing rapidly based on State Grid first-degree deployment, this paper proposed data fragmentation middleware UAPDS (UAP DB Sharding) technology. UAPDS provides multiple split solutions, such as separated of reading and writing, divided into multiple tables in a database or in multiple databases, by router, partition, shard three layer split mode, the data cut into different databases or tables according to certain rules, effectively relieve the pressure on the database, improved the efficiency of data access, achieved better in the practical application effect.