EMC claims to 'Boost' Data Domain deduplication speed by 50 percent

Boost add-on, which company says can reduce backup traffic on the LAN by 80 percent, will be integrated with EMC's NetWorker software later this year

EMC announced at its annual user conference here on Tuesday a new software add-on that it claims increases the performance of its Data Domain appliance by an average of 50 percent.

Data Domain's new Boost software, which achieves the speed increase by offloading parts of its deduplication process to backup servers and thereby freeing up CPU cycles, now only works with Symantec's NetBackup and Backup Exec backup software.

[ Keep up with the latest approaches to managing information overload and staying compliant in InfoWorld's Enterprise Data Explosion newsletter. ]

However, EMC plans to integrate the add-on with its own NetWorker software in the second half of this year.

Data Explosion iGuide

While offloading "fingerprinting," or the ability to identify duplicate data, to a data center's backup server sounds counterintuitive, EMC said moving those functions to the media server means it, in turn, will send less data across the LAN to the Data Domain appliance. This reduces both the LAN bandwidth and appliance processing requirements.

The Boost feature on a Data Domain appliance can reduce backup traffic on the LAN by 80 percent to 90 percent, said PK Gupta, EMC's director of backup recovery systems for Asia, Pacific, and Japan operations. As an example, EMC said a flagship Data Domain DD880 appliance's throughput increases to 8.8TB per hour with the Boost feature, from 5.4TB when the appliance was first introduced.

"The integration of backup software with deduplication storage is not just about enhancing performance; it's about increasing functionality to enable a more streamlined and sophisticated user experience," said Laura DuBois, a program director for storage software at research firm IDC.

PK Gupta, EMC's director of backup recovery systems for Asia, Pacific, and Japan operations, admitted that until now, Data Domain's deduplication appliance has been better integrated with competing backup software, "but that story is changing."

Gupta said Boost would be available for EMC's own Networker backup software later this year affording functions such as auto discovery, auto configuration, monitoring, and reporting, which "are much better integrated with NetWorker than NetBackup."

Once integrated with NetWorker, the backup software will also be able to manage backups that are deduplicated by EMC's Avamar deduplication software, which is already integrated with NetWorker.

Prior to purchasing Data Domain last year, EMC already had a deduplication product that was integrated with NetWorker. EMC's Avamar software deduplication product, which it purchased in 2006.

Rod Matthews, senior director of business development for backup and recovery systems, said that Avamar and Data Domain products have distinct and different use cases and being able to manage both through NetWorker will reduce management requirements.

"One optimizes use cases that are bandwidth constrained or focused on an end-point, which is Avamar scenario. Then there's the Data Domain scenario, which is optimized for a data center target environment, such as database backup, mainframe backup, AS400s," he said. "The world needs both, and we're going to offer both."

Matthews said EMC will continue to integrate Data Domain's deduplication capabilities with other products in the future. "The long-term strategic vision is you've got one software stack on the front end, with the Avamar and NetWorker stack are one client you get today," Matthews said. "The next step is how do we integrate the backend where there's one storage device that it all writes to. That's a longer-term project."

"Using Data Domain deduplication with various archiving workloads in addition to things like Centera and Celerra is another place you'll see tighter integration happening," he said.

Lucas Mearian covers storage, disaster recovery, and business continuity, financial services infrastructure and health care IT for Computerworld. Follow Lucas on Twitter at @lucasmearian or subscribe to Lucas's RSS feed. His e-mail address is lmearian@computerworld.com .

Read more about storage software in Computerworld's Storage Software Knowledge Center.

This story, "EMC claims to 'Boost' Data Domain deduplication speed by 50 percent" was originally published by Computerworld.