【摘要】 Hadoop作为开源的云计算框架已得到广泛应用,而对其分布式文件系统(Hadoop Distributed File System,HDFS)的管理却一直是项繁琐的工作,如何解决HDFS集群资源的管理问题已成为焦点,也被Apache提升为高层次Hadoop子项目。结合Java Swing技术和HDFS API实现的HDFS资源管理器能有效的缓解此问题。本文首先介绍了HDFS和Windows资源管理器,进而提出HDFS资源管理器的设计目标,并通过调研当前国内外对于HDFS资源管理器的研究进展,提出了系统管理HDFS集群资源的整体架构,重点描述了系统功能设计以及主要功能的实现原理,最后简要介绍了系统开发测试环境和优缺点。该系统还可实现对HDFS集群资源的远程可视化操作及管理。更多还原

【Abstract】 Hadoop, as an open source cloud computing framework, has been widely used, but it is always a tedious work for the management of its Distributed File System(HDFS). How to solve the management issues of HDFS cluster resource has became the focus and has also been promoted to the high-level Apache Hadoop subprojects. Combining with Java Swing technology and HDFS API, this HDFS Explorer can effectively alleviate this problem. This paper introduces HDFS and Windows Explorer firstly, and then leads to the design goal of the HDFS Explorer. Through review of the research progress of HDFS Explorer, we proposed a overall framework to manage HDFS cluster resource, and focused on the system design and the realization principle of main functions. Finally we make a brief introduction of testing environment for the system development, and talk about its advantages and disadvantages. This system can also be used to realize remote visual operation and management of the HDFS cluster resources.更多还原