应用Hadoop提升数据库审计系统检索性能的研究

Study on Applying Hadoop to Enhance the Retrieval Performance of Database Audit Systems

鉴于单节点数据库审计系统检索性能低下的现状,探讨应用Hadoop伪分布模式和HBase列存储模型重构数据库审计系统的检索存储体系,重点研究HDFS存储机制、Map Reduce运算框架和HBase数据模型三者的集成,以提升数据库审计系统实时检索和综合分析的性能.重构方案有效提升了检索性能,但鉴于数据的高可靠性和大体积,提出结合生产现状应用Hadoop和HBase分布式集群的展望.

Based on the present situation of lower retrieval performance of the single node database audit system, this paper carries out research of application of the Hadoop pseudo distribution model and the HBase column storage model to reconstruct the database audit system of the single node deployment for enhancing the retrieval performance. This paper focuses on the study of integrating HDFS storage mechanism, Map Reduce computing framework and HBase data model to improve the performance of real-time retrieval and comprehensive analysis of database audit system. Reconstruction scheme can effectively improve the retrieval performance, but in view of the high reliability and large volume of the data, the application of Hadoop and HBase distributed cluster is proposed in combination with the production status.