Hewlett-Packard is looking to boost a data deduplication acceleration technique originated by EMC. HP today extended its StoreOnce data deduplication platform with the addition of StoreOnce Catalyst software that enables deduplication at the application and backup servers.
Catalyst works with HP’s StoreOnce B6200 disk backup target appliance by deduping data on application and backup servers before moving the data to the disk system. Catalyst, which was launched last November, is similar to EMC’s DD Boost software that EMC introduced in 2010 for its Data Domain backup targets.
The B6200 is HP’s largest StoreOnce device, scaling from a two-node 48 TB model to 768 TB of raw capacity. Catalyst can run on the B6200 target or the application server. HP claims a full-capacity B6200 can back up data at 100 TB per hour and restore at 40 TB per hour with Catalyst. EMC says its flagship Data Domain DD990 can back up at 31 TB per hour with DD Boost. Of course, all dedupe speeds also depend on the type of application data and other factors such as network connectivity.
On the application side, Catalyst supports HP’s Data Protector 7 and Symantec NetBackup. HP plans to add support for Symantec Backup Exec later this year, according to Sean Kinney, director of product marketing for HP storage. DD Boost supports EMC’s Avamar and NetWorker, as well as NetBackup, Backup Exec, Quest Software vRanger and Oracle RMAN.
EMC is the data deduplication leader with its Data Domain backup target platform and Avamar client-side dedupe software. Although HP was late to the dedupe market, Kinney claims StoreOnce is superior to EMC’s dedupe because Avamar and Data Domain use different algorithms.
“They use individual point products that weren’t designed to work together from the ground up,” he said. “If you want to move across those environments, you have to rehydrate data again.”
Enterprise Strategy Group senior analyst Jason Buffington said Catalyst improves StoreOnce’s dedupe by extending it to the source where data is created. He said Catalyst makes the production server aware of what data is already is already on the dedupe target, so it transmits less data and improves the speed of backups.
“I characterize it as good, better, and best dedupe,” he said. “Good dedupe happens just by having it because it optimizes storage. Better dedupe is when the backup server is dedupe-aware so it doesn’t send data to the storage device that is already on the storage device. The best dedupe is having that level of discernment at the production server. You want to get the intelligence of dedupe closer to the product workload, and that’s what Calayst does.”
Catalyst’s pricing starts at $37,500 for two B6200 nodes.
HP also launched the latest version of its Data Protector backup software today, integrating it with features from Autonomy, the information management company that HP bought for $10.3 billion last year. The new Data Protector 7 supports the HP Autonomy Cloud, a private cloud built on the HP Converged Cloud. Data Protector 7 is also integrated with Autonomy IDOL (intelligent data operating layer), which indexes and protects unstructured data, including social media, video, images, audio and email.