中国科技论文在线中文稿件模板-HBase 加载数据的研究.pdfVIP

中国科技论文在线 HBase 加载数据的研究 冯晓普* （北京邮电大学计算机学院，北京 100876） 5 摘要：因特网在今天已经取得了巨大的进步。如今每天都会有海量的数据被上传到互联网上。 随着数据规模越来越大，大量业务场景开始考虑数据存储水平扩展，使得存储服务可以增加、 删除，而目前的关系型数据库更专注于一台机器。海量数据量存储成为瓶颈，单台机器无法 负载大量数据。HBase 是从Hadoop 中分离出来的Apache 顶级开源项目。由于它很好地用 10 Java 实现了Google 的BigTable 系统大部分特性，因此在数据量猛增的今天非常受到欢迎。 如何向HBase 中加载海量数据是使用HBase 的第一步，本文针对HBase 自带的几种不同的 加载数据方法进行了研究与性能对比，并实现了自定义并行加载数据方法，实验表明其具有 良好的效率。 关键词：HBase ；Hadoop ；Apache ；加载数据 15 中图分类号：TP392 The study of loading data into HBase Feng Xiaopu (School of Computer Science,Beijing University of Posts and Telecommunications, Beijing 20 100879) Abstract: Today Internet has made tremendous progress. Now every day there are vast amounts of data to be uploaded to the Internet. With the increasing scale of data, a large number of business scenarios began to consider the level of data storage expansion, storage services can add and delete.But the current relational database focuses more on a machine. The massive amount of data 25 storage becomes a bottleneck, a single machine can not load large amounts of data. HBase is a top Apache open source projects that separated from Hadoop. As it has most of the features of Googles BigTable system and is implemented in Java, therefore it is very popular in the days full of data. How to load huge amounts of data into HBase is the first step to use HBase. As HBase has several different methods to load data,this article made a exhaustive study of them and made a 30 performance testing of them.Also,this article achieved the custom loading data, and exper