| Cloud computing draws a lot of attention in recent years.Cloud storage is a popular applications of cloud computing,it has excellent performance in huge amount of data storage,this is far superior to traditional storage system.The development of traditional storage system has been difficult to keep up with t the pace of demand growth data.so it has become a hot issue quickly.Cloud storage is a service which designed to meet the user's storage needs,user's data will be stored on a server cluster,user can access the data stored in the system anywhere or what kind of terminal as long as the cloud storage system can be connected.It provides the user with a unlimited capacity mobile hard disk,and the security of the data is much better than the mobile hard disk.Hadoop is Apache senior project that supports distributed applications which involves huge amount of data and thousands of nodes.It is a software framework,free and open source.It is being developed by global volunteers,implemented in Java.HDFS is one of Hadoop's subprojects,it is a distributed file system which provides the high performance of Hadoop by giving high throughput access to data,inspired by Google File System.Although Hadoop is widely used,it has some defects which affect its performance,one of them is the small files problem.Hadoop Archives and sequence files are two existing solutions.But they still have their shortcomings.This paper propose a solution which can make the HDFS has better performance in handing small files,because it can reduce NameNode memory consumption and improve access efficiency.In addition,this paper build a cloud storage system based on HDFS,this system designs and implements some fundamental functions such as register,login,upload file,browse file,download file,share file and so on.Although the function is still not complete and detailed,it can deepen the understanding of cloud storage system,this is of great signifycance to the further study of extension application which based on HDFS. |
PDF Full Text Request