NetApp Adds Data De-dupe To Cut Storage Requirements

While working on a recent article on data de-duplication, I wondered why no NAS vendor had adopted the technology. When you consider the number of files sent as e-mail attachments to 50 users and saved by five of them to their home directories, just file-level, single-instance storage should save most organizations 20% or so of their file server disk space.
NetApp today announced A-SIS (Advanced Single Instance Storage) that, for the first time, brings data de-duplication technology to primary storage. A-SIS goes further, doing de-duplication on 4K blocks through NetApp's WAFL file system. For stored files this means it can determine that 90 percent of the PowerPoint presentations with only a different title slide for different customers are the same data and store just one copy. Since WAFL stores data spread across the volume's space already, there should be a minimal performance hit reading de-duped files. Since iSCSI and Fibre Channel volumes on a Filer are stored as WAFL volumes, A-SIS can also de-dupe these LUNs created for block level access. De-duping runs as a post-process, identifying and eliminating duplicate blocks in the background or on a schedule. Contrary to their usual practice of "nickel and diming" customers to death, NetApp is giving away A-SIS to users of their NearStore products, which are frequently used as backup targets. FAS filer users can license A-SIS for $3000, which in the world of enterprise storage is a bargain.
Howard Marks
NWC Contributing Editor