The Technical Research Of Double Fingerprint Check And Multithreading Transmission In Cloud Backup

With the improvement in the level of computer information, computer has been widely used in all fields of society, which results in data's growing explosively in geometric progression. In order to keep the security of data, as a new technology, Cloud backup provides a reliable data protection solution, while the speed of backup has been the key factor preventing the popularization of it.The paper proposed a solution of optimizing the speed of data backup, based on the study of deduplication and data transmission in the Cloud Backup. In the traditional deduplication, the system cuts data objects into blocks and computes fingerprint for each blocks, in the basis of which the system can check whether the blocks are repeated or not. However because the collision of fingerprints, which means that different blocks may compute a same fingerprint, high-strength calculation is used to obtain fingerprint with a large numbers of bits so as to avoid collision. In the mean time, fingerprint calculation brought by a large amount of data can't be ignored. The Double fingerprint check can avoid a large number of fingerprint calculations. First, eliminate most of non-duplicate data blocks through low computational fingerprints. Secondly, use high computational fingerprints to process the remained data blocks. As a smaller proportion of duplicate data with localized features, this method can reduce the computation effectively. Moreover, the network data transmission can't take full advantage of the bandwidth for the characteristics of the data block in deduplication. Using multithreading which means that transferring data through multiple network sockets can take full advantage of the bandwidth, and it can improve data transmission speed.The performance of double fingerprint check and multithreading transmission are tested by the BCloud online backup software, the results of experiment show that both methods can achieve the purpose of reducing the computation and improving data transmission efficiency, and it also optimizes the overall performance of data backup.