Data Domain Adds Retention Enforcement - Deduplication, It's Not Just For Backup AnymoreData Domain Adds Retention Enforcement - Deduplication, It's Not Just For Backup Anymore
In a further attempt to position its deduplicating NAS appliances as general purpose data repositories, Data Domain has added date retention enforcement as an optional feature. This follows naturally from the redesign of the file system last year to support a large number of small files as well as the small number of large files typical of a backup target.
In a further attempt to position its deduplicating NAS appliances as general purpose data repositories, Data Domain has added date retention enforcement as an optional feature. This follows naturally from the redesign of the file system last year to support a large number of small files as well as the small number of large files typical of a backup target.At least in this initial release, Data Domain's solution allows the system administrator to override retention periods, so it's not suitable for Wall Street, pharmaceuticals, and other heavily regulated industries where total immutability is required.
Like several other locked NAS vendors, Data Domain has adopted the mechanism of using the last accessed date field to specify the retention period, an API first used by NetApp's SnapLock option. An application, or user, sets the last access date to the end of a file or folder's retention period and then flips the read-only flag on the file. From that point on, the read-only flag can only be reset by the system administrator before the date specified as last accessed date. Since this API already is in wide use, most archiving applications from Disk Extender and Enterprise Vault to F5 and Attune's file virtualization solutions already support it.
Using a single reduplicating repository for archives and backups should result in lower disk usage, as there will be some data commonality like files that appear in the production file servers and the e-mail archives. This is especially true for a system like Data Domain's that uses variable length blocks and standard data compression rather than context-based deduplication that can be very application specific.
About the Author
You May Also Like