EMC wants to extend Data Domain's domination of the data deduplication backup target market to archiving. The new version of Data Domain's operating system has been enhanced to work better with archive data and supports more archiving applications.
Data Domain's 5.3 operating system also includes wider application support for DD Boost and a new interface to manage multiple appliances.
EMC also released SourceOne 7. The latest version of the archiving software adds file archiving, the ability to index file system data "in the wild" -- regardless of whether the files are stored in the archive -- and improved auditing and reporting.
Data Domain's file system has been upgraded to improve the performance of ingesting large numbers of small files, which is more common for archiving apps than backup apps. The latest version can migrate data from other archive systems three times as fast as the previous version, according to Peter Smails, EMC director of product marketing.
"File ingesting for backup is materially different than for archiving," Smails said. "We needed to enhance our capabilities around smaller file support. Handling a large number of smaller files is different than handling a small number of large files, which is your typical backup workload."
Smails said the main focus of the upgrade is the ability to archive data on the backup target, although backup and archiving are separate processes.
"You still need different applications to do backup and archiving, but we see benefits of consolidating backup and archiving workloads on the back end," Smails said. "Archiving is a logical adjacent market to backup."
Arun Taneja, consulting analyst of the Taneja Group, said EMC is taking a different approach to backup and archiving than most vendors.
"EMC has decided the repository for backup and archiving will be the same," he said. "For most vendors, those worlds are separate. Previously we had to use different file systems to handle a small number of large files than to handle a large number of small files. But EMC has found a way to make the Data Domain file system manage both of those things. It simplifies things if there's one system and the user doesn't have to know the boundaries."
Smails said EMC still considers its object-based Centera appliance a good fit for compliance-focused applications but positions Data Domain for companies who want to consolidate backup and archiving. Centera has been around for 10 years and EMC has not upgraded the platform much recently, leading to expectations in the industry that it is being phased out.
"I think Centera has run its course," Taneja said. "It has served its purpose well, but it's time to let the poor guy find its pasture. I think strategically Data Domain will be EMC's archiving platform, even for compliance."
Yet not all EMC customers are ready to give up on Centera.
Sean O'Mahoney, managing director of technology services for Republic Bank and Trust Company in Louisville, Ky., said he plans to continue to use Centera for archiving while backing up to Data Domain. Republic also uses SourceOne software for e-mail archiving. The bank has Data Domain DD670 appliances at its main and second data centers, and also has Centera boxes in both locations.
O'Mahoney said he uses Data Domain for what he calls "semi-archiving" of backup sets that he retains long-term, but he is happy with Centera as his main archiving platform.
"We archive everything on Centera. It gives us replication and is a viable archive target for SourceOne," he said. "We feel comfortable with Centera's content-addressed storage [CAS] instead of straight file-based storage, and it manages its own replication."
Data Domain added support for IBM Optim and EMC Documentum archiving apps.
Data Domain OS 5.3 also includes more support for DD Boost, which speeds backups by offloading part of the data deduplication process to the backup server. The added support includes:
- Multiplexing in Oracle RMAN. EMC first added DD Boost for RMAN a year ago, but that was only for single streams and had less of a performance benefit.
- Fibre Channel backups to go with previous DD Boost support for NAS-based backups. EMC claims DD Boost provides 50% faster backups over FC than backups without DD Boost. Most Data Domain customers use NAS for backups, but Smails said many larger customers are now using FC.
- Support for Symantec NetBackup's auto image replication (AIR) feature, which aligns the backup image and metadata more closely during replication.
- Support for managed replication in Greenplum data analytics appliances.
EMC's new Data Domain Management Center (DDMC) allows customers to manage all Data Domain devices from one interface. Previously, each box had to be managed separately.