FalconStor has announced its Single Instance Repository (SIR), a new product to de-duplicate stored data - at both file and block level.
It is for its virtual tape library (VTL for tape data held on disk) and can increase effective capacity by twelve times. Where 30 days worth of recovery points can be kept on a VTL now, with the de-dupe product up to a year's worth could be kept.
Network bandwidth needed to access a remote VTL can also be reduced.
FalconStor SIR utilises a Redundant Data Elimination (RDE) Engine to scan the VTL for duplicate files. Only unique data is then stored in the repository.
The resulting “virtual” tape, also known as an index tape (i-Tape) is stored back in the VTL at a percentage that is a tiny fraction of the size of the original virtual tape. Each virtual tape is metadata linked to the repository for instant recovery when needed.
In effect this is a new form of synthetic backup with restored files being constructed from stored disk blocks, pointers and meta data.
Naturally backed-up data that has already been compressed, such as jpeg files, can't be compressed any more. The de-duped data can also be encrypted for added security.
FalconStor is certifying the product with backup application vendors. Files can be retrieved from the de-duped VTL repository just by name and version and time/date without having to go through a backup application's restore process. Users can do this themselves. With disk-to-disk backup (which is what a VTL is essentially) being so fast and with restoration not needing backup software we have to ask will backup software really be needed at all in the future?
Possibly we'll wonder in future why so much collective IT effort was expended on backing up duplicated data, causing enormous wasted expense both of disk capacity and network bandwidth.