摘要
HBase是一个面向列的非关系型的开源分布式存储系统,它基于Hadoop HDFS文件存储系统,使用MapReduce来处理海量数据,利用Zookeeper作为协同服务,它使用简单的键值对映像关系为超大规模和高并发的海量数据实时响应系统提供一个很好的解决方案。对HBase的体系结构、数据模型、MapReduce算法设计等几个方面进行详细阐述,并对HBase的未来发展做出展望。
HBase is a column-oriented non-relational open source distributed storage system, which is based on the Hadoop HDFS file storage sys- tem, it uses MapReduee to process vast amounts of data, and Zookeeper as a collaborative service, it uses a simple Key-Value relation- ship for the massive data size and it is a good solution for high concurrent real-time responded system. Describes the HBase architecture, data model in details, and the MapReduce algorithm design, and makes the outlook for the future development of HBase.
基金
武汉科技大学教学研究项目(No.2006100)
