面向在线日志云的海量数据分级归档机制  

Hierarchical archiving mechanism of big data for online log cloud

作　　者：杨锋英[1] 刘会超[1,2] 

机构地区：[1]黄淮学院信息工程学院,河南驻马店463000 [2]武汉大学计算机学院,武汉430072

出　　处：《信息技术》2014年第8期45-48,共4页Information Technology

基　　金：河南省科技攻关计划项目(122102310474);驻马店市科技发展计划项目(11314)

摘　　要：传统日志系统和基于Hadoop的离线日志系统在分析海量日志时都具有时延长、效率低等弊端。为构建新的基于Hadoop的在线日志系统,提出了新的数据分级归档机制,用以管理海量日志数据并形成分级优化的存储文件结构。实验结果表明该机制具有良好的可扩展性,可以有效存储海量日志数据;同时,它还有助于加快系统的数据处理速度,缩短系统的处理响应时间,使其满足在线处理的要求。The traditional and the offline Hadoop-based log system all have the drawbacks of longerprocessing delay and lower efficiency when processing the massive log data. In order to build the newonline Hadoop-based log system, this paper introduces a new hierarchical data archiving mechanism,which used to manage the large amounts of log data and form hierarchical and optimized storage filestructures. The experimental results show that the mechanism has good scalability and can effectivelystore massive amounts of log data. Meanwhile, it also helps to accelerate the data processing speed,shorten the system response time, which can meet the requirements of online processing.

关 键 词：云计算 分级归档 海量数据 日志云 在线处理 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]