基于Hadoop的自动售票日志分析系统设计  

TVM Log Analysis System based on Hadoop

作　　者：王斌[1] 李超[2] 蒋秋华[2] 

机构地区：[1]中国铁道科学研究院研究生部,北京100081 [2]中国铁道科学研究院电子计算技术研究所,北京100081

出　　处：《铁路计算机应用》2014年第7期20-23,共4页Railway Computer Application

摘　　要：通过对自动售票系统日志数据分析,不仅能了解系统的运行状况,还能在日常维护中更容易地发现问题,降低维护成本,提高维护的效率。针对这一目的,本文基于Hadoop框架设计一个自动售票系统的日志分析系统,搭建Hadoop集群环境,并对处于不同自动售票应用服务器上的日志进行收集,针对自动售票系统日志的特点,编写Map/Reduce算法,对收集到的日志进行分析处理,使其结果满足需求,同时,验证基于Hadoop的自动售票日志分析系统的有效性及可行性。TVMs could produce a large number of log data which included much valuable information.From these information,we could know the status of our system and maintain the system more efficiently.In all ways of data processing,Hadoop was an open source framework which was used widely in large data sets processing.For this purpose,this paper designed a log analysis system of automatic ticketing system based on Hadoop,in this way,we could collect and analyze the log data of TVM and make the result meet our demand by Map/Reduce Algorithm.Meanwhile it was verified that the System was effective and feasible.

关 键 词：HADOOP 自动售票机 分布计算 日志分析 

分 类 号：U293.22[交通运输工程—交通运输规划与管理] TP39[交通运输工程—道路与铁道工程]