EMC has announced the Data Domain Global Deduplication Array (GDA), which the company claims is the industry's highest performance inline deduplication storage system. The system consists of two Data Domain DD880 controllers combined into a single global name space, says Steve Jackson, senior director of product marketing for EMC's backup recovery systems division. "To the backup application, it looks like one system," he says.
At first release, the EMC GDA will work with Symantec Net Backup and Backup Exec through OpenStorage, and in the second half of the year, it will support EMC NetWorker, Jackson says. While it might seem odd that Symantec backup software is supported before EMC's, this is because Data Domain itself, which had supported Symantec software, has only been a part of EMC for the past nine months, he says. "We'll catch NetWorker up later in the year." The combined system offers throughput of 12.8TB/hour, and a capacity of 280TB, meaning that with a 50-1 deduplication ratio, the system supports up to 14.2PB of logical capacity, Jackson says.
The GDA offers a great deal of scalability both physically and logically, as well as its throughput, says Brian Babineau, senior consulting analyst for Enterprise Strategy Group Inc. "Customers can protect more data to a single system within their allotted backup windows," he says. "The consolidation then facilitates efficient disaster recovery, because it is easier to copy everyone from one system to another as opposed to copying many systems to many systems." Part of how the GDA achieves that scalability is to execute a portion of the deduplication on the back-up media server, as opposed to doing it all on the system, Babineau says. However, it is not a clustered architecture, he says. "If one of the individual systems goes down, the entire solution is unavailable." On the other hand, it is still likely to be more available than tape systems, and users are more likely to put performance and capacity ahead of availability in their evaluations, he says.
The system supports up to 270 originators, meaning it can back up up to 270 remote systems at once, says Jackson. However, compared to a tape library, it reduces complexity, he says. "It takes a lot of work on the part of the administrator to figure out what backup jobs go where," he says. With the EMC GDA, "you point all your back-up jobs at one system and we figure out the best place for the backup." The system also offers bi-directional replication, meaning it can push copies of data out to multiple destinations, such as for disaster recovery or development testing, he says.
EMC also announced that the DD880 itself has had its performance increased from 5.4TB/hr to 12.8TB/hr, and its capacity doubled, Jackson says. In addition, EMC is now offering a 128- and 256-bit encryption option that encrypts data before it is written to disk, improving security and making it easier to pass security audits, he says. The products will be available before July 1 starting at $800,150.