EMC® recently announced new Data Domain systems, the Data Domain DD890 and DD860 appliances, the new Data Domain Global Deduplication Array as well as Data Domain Archiver, the industry’s first disk based system for long-term retention of backup and archive data.
Data Domain DD860, DD890, Global Deduplication Array and other enhancements
EMC has, once again increased throughput and capacity on their high-end appliances. Specifically,
- The new DD890 takes over from their former high-end system, the DD880 with close to double the throughput (14.7TB/hr with DD Boost or 8.1TB/hr with VTL) and double system capacity at 384TB raw (285TB usable).
- The new DD860 appliance replaces for their old midrange DD690 with more than doubled throughput (9.8TB/hr with DD Boost or 5.1TB/hr with VTL) while quadrupling capacity with up to 192TB raw (142TB usable).
- The new GDA now consists of two DD890’s and adds support for Data Domain VTL software in addition to DD Boost with more than double the throughput (26.3 TB/hr with DD Boost or 10.7 TB/hr with VTL).
Also, Data Domain now supports IBM i (AS/400 on IBM Power System) backup services across their product line. This is a new software feature that can be enabled for any Data Domain appliance.
Data Domain Archiver
As EMC says, archive is a journey not a destination. Most of today’s archive products consist of point solutions, not generally applicable outside a specific application domain. Historically, if one needed long-term retention of other data, one kept full backups around longer and prior to data deduplication, these were almost always tape volumes. The new DD Archiver solution extends Data Domain architecture with an internal tiering approach that enables long-term retention of data on deduplicated disk without requiring the use of tape.
The DD Archiver supports up to 768 TB of raw capacity that can be configured as either a short-term active tier (for operational backup and recovery) or long-term archive units for file data. Up to 128 TB of raw capacity in a DD Archiver can be used for short-term storage with the remainder dedicated to long-term retention.
To support long-term preservation, the DD Archiver will automatically and transparently move data from the active tier to the archive tier as data ages by policy. When an archive unit or disk shelf fills up, the unit is then “sealed”, by giving it all the deduplication and file system metadata needed to be completely recoverable by itself, without requiring other disk shelves. Not unlike a full backup tape volume set consisting of a self-contained data backup.
Backup or archive software will take care of file expiration and trigger the deletion of expired files on DD Archiver over time. For example, to provide a 7-year archive, the backup application would delete 7-year old backup files on DD Archiver as they age, over time.
Furthermore, the DD Archiver can be a target for data replication of other Data Domain appliances. In this way, archiving activity can be consolidated to a central DD Archiver. Alternatively, one DD Archiver can be replicated to another DD Archiver to provide additional offsite data protection.
Today, the DD Archiver supports NAS protocols and DD Boost data access but EMC will likely support VTL in the near future.
If I didn’t know better, I would say that EMC didn’t care for tape. In the background information EMC provided a chart indicating that tape automation’s share of external IT spend has been shrinking over the last 5-years and is now below 7% (’09). Furthermore, EMC indicated that one of the last remaining holdouts for tape use was long-term archive.
Presumably, tape will not give up on this arena very easily. The economics and usability for data deduplication are not quite as clear in archive as they are for normal (short-term retention) backup operations. On the other hand, tape infrastructure can be an expensive proposition for archive use alone.
A PDF version of this can be found atEMC 2011 Jan 18 New BRS Data Domain systems and Archiver
Silverton Consulting, Inc. is a Storage, Strategy & Systems consulting services company, based in the USA offering products and services to the data storage community.