Research On Distributed File System Based On Dynamic Multiple Center On Cloud Stroage

In the era of big data, cloud storage technology is the basis of data mining, dataanalysis and research, which becomes a hot issue. As the core technology of cloud storage,the distributed file system has a new research value. The article makes a comprehensiveanalysis of distributed file systems in the domestic and overseas, and points out thedeficiency in the metadata management. In order to ensure the system’s reliability andimprove the system’s throughput, it raises a distributed file system based on multicenter,and conducts the in-depth study of the key technology of the model involved in.Firstly, in view of the bottleneck of the single center node in the present distributedfile system, it designs a frame model based on a dynamic multicenter distributed filesystem. which increases the reliability and the throughput of the system. The model usesthe method of N heavy Hash and layered combination to organize the metadata in memorystorage, uses the log strategy to backup the metadata, and proposes a fault switchingscheme based on three cycle chain.Secondly, it carries on the analysis and research for the load balancing problem of thenew model, and proposes a dynamic adaptive load balancing strategy, which decouples theload balancer and client to improve the system’s reliability, uses the mechanism ofpredicting to reduce the overhead of system migration tasks,Finally, it also analyzes and studies for the problem of data consistency, and putsforward a synchronization scheme based on improved Paxos algorithm, by combining thesynchronization scheme of periodic and random, which can reduce the size of“inconsistency window” time of the system model at the largest extent. Meanwhile, itvalidates the accuracy of the synchronization scheme through the experiment ofengineering for multicenter node.