Design And Implementation Of Cloud Security Storage System Based On Hadoop

Cloud computing is the development of distributed computing, parallel computing and grid computing, to bring about a fundamental change in human life, production methods and business models, is one of the hot spots in current society. Cloud storage is a new concept extends from the cloud computing, and it is a system that through the application cluster, grid technology and distributed file systems and other functions, puts large number of different types of network storage devices set by the application software to work together to provide external data storage and service access function. With the wide application of cloud computing technology and cloud storage technology, security problems have emerged, and have become the key factor hindering the development of cloud computing and cloud storage.Hadoop is an open source distributed computing platform under the Apache Software Foundation, provides developers with a distributed system infrastructure, users can develop distributed applications in case of lack of understanding of the underlying details of distributed systems. Distributed file system HDFS as the core component of Hadoop, has powerful storage capability, suitable for cloud storage system. However there are some design flaws, it is still evolving and has improve and perfect space.This thesis studies the HDFS multi-replication technology and authentication technology, proposes improved methods for the existing problems. Studies the security issues of data storage in the environment of cloud computing based on Hadoop, designs a cloud security storage system and implements a system prototype. The main work is as follows:(1) Analyzes the low storage efficiency and easy to appear imbalance problem, proposes a distributed storage method based on information dispersal algorithm and designs an improved replication storage strategy, simulation results show the effectiveness of the method.(2) Points out the HDFS authentication is easy to impersonate a user and its other problems, designs-a secure and effective one-time password authentication scheme, and uses this authentication scheme combined with HDFS authentication to authenticate the user, improves the security of HDFS.(3) Using JSP technology combined with Hadoop API has developed a cloud security storage system based on Hadoop, visualization of HDFS interface, implements file transfer encryption, access control and security sharing, overcomes the HDFS plaintext transmission and storage of data security risks, improves the security of user data.The inovations of the thesis includes:(1) Proposes a distributed storage strategy based on information dispersal algorithm and the improved replication storage strategy can improve the HDFS’s load balancing capabilities.(2) Designs a secure and effective one-time password authentication scheme, uses this authentication scheme combined with HDFS authentication to improve the security of HDFS.(3) Designs a cloud security storage system based on Hadoop, implements the system prototype.