The Optimization And Implementation Of Cloud Storage System Based On HDFS

Cloud computing has been a research focus currently. As a derivative of cloud computing,Cloud storage is also one of the most popular study fields in China. As the open sourceimplementation of the Google File System(GFS), Hadoop file system HDFS actually becomethe standard of the study on cloud computing and cloud storage and the implementation ofcloud applications and cloud services. However, HDFS architecture has a number ofshortcoming, typically including lack of support for small files, performance bottlenecksbecause of the single NameNode. How to solve these problems, is one of the hotspots of thecurrent study.Based on the study of HDFS, the paper proposes some appropriate solutions. For smallfile problem, this paper proposes a way to introduce user metadata space to merge the smallfiles for large files in the HDFS storage; for the HDFS NameNode single performancebottleneck, this article proposes a NameNode solutions based on MongoDB. Experimentalresults show that this proposal, not only expanding the HDFS namespace clusters, andconcurrent read and write speeds of HDFS.In addition, this article also set up a cloud storage system on the basis of the HDFSarchitecture, to achieve file upload, download, share, browse, and other functions. At thesame time, the system can also monitor current HDFS cluster, monitor capacity information,including cluster information, load information, CPU usage information for a single node.Implementation of cloud storage system has a great exploring and guiding significance forapplications based on HDFS.