Research On Cloud Data Integrity Verification And Data Recovery

Cloud computing is a new method of shared infrastructure via the internet, which puts a lot of system resources together and provides users with a series of more convenient services with virtualization technology. Cloud storage system is mainly responsible for data storage and data management. Although cloud storage system brings a lot of convenience, cloud storage security problems are also increasingly obvious, including the data integrity. Users must have an effective solution to check their cloud data. When data corruption has been detected, users must know the location of the error(s) and further recover from failure. In addition, in order to meet the needs of users to dynamically modify the file, the solution needs to support both integrity detection and dynamic data operation.Aiming at integrity detection and data recovery, we will finish the following work.(1)This thesis first introduces the research of cloud data integrity, summarizes the advantages and disadvantages of the existing data integrity verification method. Finally, mainly introduces the RS code, which is an erasure code.(2)This thesis proposes a data integrity verification model based on RS code for distributed cloud storage(DIMRS). It supports unlimited data integrity verification. It can detect data corruptions with low communication overhead and overwhelming probability.(3)When corruptions occurs, this thesis use an error localization technology with the pre-computed verification tokens to locate where the potential data error(s) lies in.(4)In order to recover from failure, this thesis proposes an efficient data recovery algorithm. DIMRS model has strong ability to resist corruptions. In the face of large area and high frequency data corruptions, it still can recover the original file with low communication overhead and overwhelming probability. The algorithm greatly provides security of cloud data integrity.(5)The model not only supports static cloud data integrity verification, but also supports dynamic cloud data integrity verification. This thesis implemented dynamic data update, delete, append, and insert operations efficiently.