A Dynamic Data Fault-Tolerance Mechanism for Cloud Storage

With the increasing of data in the network, and mainly using the fault-tolerance of replication, the Cloud Storage will be hard to satisfied requirements of storing all data. To improve the data fault-tolerance of cloud storage, we propose a dynamic data fault-tolerance for cloud storage (DDFMCS) that dynamically determines the data fault tolerance mechanisms by the file access frequency ratio stored in the file access frequency table, the number of file fault-tolerance conversions and the time of files stored in the system. We implement DDFMCS based on Hadoop and conduct some experiments. Experimental result shows that DDFMCS can increase the utilization of cloud storage space and improve data access performance.