基于Postgres-XL 的数据管理优化技术研究-《计算机技术与发展》

文章信息/Info

Title:: Study on Optimization Technology of Data Management Based on Postgres-XL

Author(s):: AI Li-rong; LI Kai; School of Computer Science,Northwestern Polytechnical University,Xi爷an 710129,China

摘要:: 面对城市中快速增长的数据量,单机数据库的性能明显应对不足。针对单机数据性能低下、可扩展性差以及数据库安全性等问题,利用开源分布式数据库对数据进行管理显得十分必要。 Postgres-XL 作为一个完全满足 ACID 的、开源的、可方便进行水平扩展的、多租户安全的数据库解决方案,可以有效地完成数据管理工作。为了进一步提升 Postgres-XL的效率,对 Postgres-XL 的扩展模块和存储管理模块进行了研究。首先提出利用 Postgres-XL 的表继承等特性完成表拆分,然后利用基于 SQL/ MED 的中间件把 Redis 作为 Postgres-XL 的缓存,最后在 Postgres-XL 外存管理模块中增加 SSD 缓存模块。实验结果表明,以上三种优化策略有效地降低了数据库请求的时间延迟,提升了 Postgres-XL 的效率。

Abstract:: As the amount of data in the city grows rapidly,the performance of stand-alone database is clearly deficient. In order to solve problems of low performance,poor scalability and the security of database,it is necessary to use the open-source distributed database in data management. As an open source and horizontally-scalable database that meets the requirement of ACID and is secure for multiple tenancies,Postgres-XL can efficiently complete the data management. In order to further improve the efficiency of Postgres-XL,the expansion and storage management modules of Postgres-XL are studied. Firstly we propose to utilize the table inheritance and other characteristics of Postgres-XL to complete the table split,then use Redis as the cache of Postgres-XL based on SQL/ MED middleware,and increase SSD cache modules in the external memory management module of Postgres-XL in the end. Experiments show that proposed optimization strategies can effectively reduce the latency of the database request and improve the processing efficiency of Postgres-XL.