File-match Based Light-weight Differencial Incremental Bakcup System

IT systems are prone to disruptions that result in data loss or corruption. When loss occurs, an often taken option is to restore data from the backup copies. But as data volume continues to rise, traditional data backup solutions are facing increasing number of problems ranging from time and space efficiency to restoration correctness.First, with the explosive growth of electronic data, backing up and restoring data has become time and space consuming. To combat this problem, people have resorted to incremental backup approaches in which only the increments from last backup are copied over. Two widely used schemes in this regard are Cumulative and Differential Incremental Backups (abbreviated as CI and DI respectively). Their difference is that the last backup (the reference point) in CI must be a full backup but in DI can be an incremental backup. Consequently, CI incurs higher data redundancy and DI requires orderly application of the increments. Second, data restoration often suffers the so-called file mis-match problem. When this happens, the only remedy is to change the pathname of the concerned file through manual intervention, which is tedious and prone to error.This paper presents a light-weight file-matching differential incremental backup (LFDI) scheme that overcomes the aforementioned problems. Our scheme offers hands-free restoration, and is highly efficient in both time and space cost. With these technologies, a complete backup system is developed.