| In recent years,with the rapid development of information technology,the amount of Internet data has increased exponentially.Massive data puts new challenges on storage systems.Distributed storage systems meet the need for data storage security and reliability.However,the storage server of the traditional distributed storage system is a fixed node.When the node status changes in real time,the storage policy cannot be adjusted in time to affect the efficiency of the system.Therefore,how to design an efficient distributed storage solution is a key issue when the network status changes in real time.On the other hand,in the personal field,the cloud disk is a relatively mature distributed storage product.Once it is launched,it is popular among users.However,the speed of a single cloud disk in different locations is very different even in different countries.The operator has the risk of interrupting the service,and the self-contained cloud disk client cannot reach the full speed of the broadband in most cases.How to improve the performance of the cloud disk is also a research hotspot.Network coding encodes the information and stores it,which can reduce repair bandwidth when node fails in case of maintaining the redundancy of storage system.This paper first studies a distributed storage system based on RLNC(random linear network coding)for dynamic networks.Then,the data coding and data recovery strategies are proposed by using network coding,then a practical system was implemented using the C#,and the overall performance and related parameters of the system were tested.Finally,the distributed storage system based on RLNC for cloudy architecture is studied.RLNC is used to ensure the privacy of the data.The RESTful API provided by multiple cloud disks is integrated to realize a cloud disk.The shared storage system NormalDrive works together.The actual experiments show that this system can improve the transmission performance of the network while ensuring the reliability of the cloud disk. |
PDF Full Text Request