Performance Optimization Of Data Recovery For Network Backup System Using Data De-duplication

As we know, the technology of information and network is developing very fast,human society has entered the big data times, Data is playing a more and more importantrole in our daily life. Using backup systems to backup and restore our data can be veryhelpful to data protection.The bandwidth of the link between users and backup service providers is often verylow, which results in low performance of data backup and recovery. Most backup andrecovery datasets is evolved from their previous versions, with only minor modifications,inserts or deleting, most of the files and data blocks did not change after the last backup.Academia and Industry use the source-side data de-duplication to delete the duplicate filesand data blocks, which only improves the performance of data backup.In this paper, we apply both file-level and block-level data de-duplication to datarecovery, which can certainly improve the performance of data recovery. First, we usefile-level data de-duplication to delete those duplicate files, and then we use block-leveldata de-duplication to delete those duplicate data blocks. After these two steps, the restfiles and data blocks must be transferred through the network. By deleting the redundantdata in the recovery dataset, we can significantly improve the performance of datarecovery.The tests have shown that applying the data de-duplication technology to datarecovery can significantly reduce the amount of data transferred through the networkduring the recovery process, thus saving network bandwidth resource and improving datarecovery performance.