Data recovery model based on dynamic deployment and block storage

Abstract

The reliability of data can be effectively improved based on the strategies of distributed storage and multicopy,block-based storage and dynamic recovery technology. However,the reliability of data can be further enhanced through organic integration of coding-storage,block-redundancy storage,and dynamic recovery technology.In the meantime,there presently lacks an effective model to measure the reliability of the data based on this policy.This paper proposes four dynamic data reliability assessment models(data failure probability models),while considering the storage nodes,node reliability,the number of data blocks,network status,reset rates and some other factors. The simulation analysis shows that with the specific storage strategies and network environment,this model can effectively output data reliably. Compared with traditional full( block) copy storage and coding storage,this model has higher reliability. Analysis of the normal and abnormal network environments enhances the usability of the model.