NoSQL数据库与缓存一致性研究

Research on Consistency between NoSQL Database and Cache

随着大数据技术的发展,数据库系统面临数据的高并发读写和海量数据的存储,而传统的关系型数据库不能满足这方面的需求。NoSQL因其高扩展性、高性能、容错性、高伸缩性,在大数据管理中被广泛采用。由于NoSQL不再遵循数据库ACID原则,使得分布式系统无法同时满足一致性、可用性和分区容错性,Key-Value模型作为NoSQL中被使用得最多的一种存储模型,同样存在着此类问题。基于此,以Redis为例,探讨了NoSQL中持久性存储和缓存之间保持数据一致性的方法。

With the development of big data technology,database systems are faced with high data concurrent reading and writing and massive data storage,and traditional relational databases are increasingly unable to meet this kind of demand.NoSQL is widely used in big data management because of its high-extendable,high-performance,fault tolerance,and highscalability.Because NoSQL no longer follows the database ACID principle,the distributed system can not meet the consistency,availability and partition tolerance at the same time.As one of the most used storage models in NoSQL,the above problem also exists in the Key-Value model.Taking Redis as an example,this paper discusses the method of maintaining data consistency between persistent storage and caching in NoSQL.