基于Storm的大数据指标实时计算方法

Real-Time Calculation Method of Big Data Index Based on Storm

大数据时代,面对爆发式增长的海量异构大数据,企业指标数据的实时供给能力亟待全面提升.基于流处理技术的大数据指标实时计算方法,主要由日志采集、消息管理、协调管理、实时处理等部分构成,使用Hadoop、Zookeeper、Storm、Kafka、Redis等开源软件,综合应用了数据库日志分析,流处理、内存计算等技术.本文详细论述了采用Storm技术的大数据指标实时计算方法的技术架构,实现方法及路径,同时给出了算法验证的过程和结果分析.

At the age of big data,faced the explosive growth of massive heterogeneous big data,the real-time supply capacity of enterprise index data needs to be improved comprehensively.The real-time computing method of big data index based on stream processing is composed of logs collection,message management,coordination management,realtime processing,and other parts.Not only open source software such as Hadoop,Zookeeper,Storm,Kafka,and Redis are used,but also the techniques of database log analysis,stream processing,and memory calculation are applied.This paper presents the discussion of the technical framework,implementation method and path of the real-time computing method of big data index based on Storm,and provides the algorithm verification process and result analysis.