EMC this week updated the top end of its Data Domain dedupe product line with clustering, a move that could mute competitors' claims that the storage appliances don't scale.
EMC unveiled the Data Domain Global Deduplication Array (GDA), a new product that pools two DD880 dedupe appliances across a single controller. As a result, the GDA now dedupes up to 14.2 petabytes of logical backup storage in a single system.
EMC acquired Data Domain last summer after a drawn-out bidding war with NetApp over the company.
EMC also doubled the capacity of its individual DD880 dedupe appliances to 7.1 petabytes of logical backup storage per unit. Logical backup storage refers to the amount of data that can be stored after it is deduped.
Prior to the introduction of the Data Domain GDA, competitors in the dedupe market often argued that Data Domain appliances are monolithic in nature, which means that once the maximum capacity in a single appliance is met, customers would have to invest in a completely new system rather than expanding the original system's capacity.
That argument was actually a red herring, said Shane Jackson, senior director of product marketing at EMC.
"It didn't account for the speed and accuracy of our systems," Jackson said. "It's true that, once they reached full capacity, customers needed to buy another. But plenty of customers were ready to do so, because we were focused on performance."
Keith Norbie, vice president of sales at Nexus Information Systems, a Minnetonka, Minn.-based solution provider, also said that such an argument did not reflect the benefits that the Data Domain appliances bring to customers.
"Data Domain is the high watermark of the dedupe world," Norbie said. "The limitations in processor, flash memory, disk speed and even the speed of light are constantly changing glass ceilings. And with every new version of its appliance, Data Domain pushes those ceilings higher."
Jackson said that doubling the capacity of the DD880 appliances and the ability to cluster two units together with GDA combine to increase maximum backup storage capacity by four times over previous limits.
"Even with competitors who have multiple controllers for expandability, customers still have to decide which backup jobs go to which controller," he said. "With Data Domain, even with two units on a controller, customers see it as a single backup array."
Data Domain also enhanced its storage software with the ability to encrypt the data to protect against appliance or individual disk theft, as well as new technology to improve remote replication, Jackson said.
While other vendors have made great strides in dedupe technology, for pure backup performance Data Domain is still the product line to beat, Norbie said.
For instance, NetApp has done a great job of unifying block-based data and file-based on a single appliance and adding dedupe capabilities, Norbie said. However, with NetApp, the dedupe is for primary storage, whereas Data Domain is optimized for backup storage, making the two lines serve completely different purposes.
The new DD880 and the GDA are slated to ship sometime this quarter, Jackson said. While EMC direct-sales reps have access to the Data Domain appliances, they typically work with channel partners, he said.