HDS Unifies Block, File, Object Storage Into Single Pool

HDS' new single-pool environment doesn't require fencing or some sort of clairvoyance as to where you're going to grow your data.

A big 2012 trend among Tier 1 data center system providers is the convergence of more and more functionality into fewer and smaller containers. But there is still a wide variance in the kinds of functionality that vendors are offering.

In storage, the new-generation systems have been about managing different tiers of block-level storage:

  • Tier 1, the freshest, most-used data and files for daily business, usually under 30 days old and kept on solid-state or fast hard disks;
  • Tier 2, near-line storage for data perhaps 30 to 90 days old, usually stored on SATA hard disks; and
  • Tier 3 offline archival storage, most likely housed in virtual tape libraries or digital tape.

/images/stories/HDS_logo_lg.jpgThree Data Classes All in One

Hitachi Data Systems is going one up on these conventions with its all-new Hitachi Unified Storage (HUS) system and single management framework, launched April 25. It puts block, file and object classes of data all in a single pool--something no other storage provider currently does.

"Up to now we've been managing all the different tiers of block layers, and we recently integrated in the file storage. Now we're integrating in the content/object side as well as the compute side," HDS Vice President of Software Sean Moser told eWEEK.

"The traditional definition for unified management has focused on managing file and block on these general-purpose lower-end storage platforms. This all changes today as we bring to market the first enterprise-class unified storage solution."

HDS' new single-pool environment "doesn't require fencing, or some sort of clairvoyance as to where you're going to grow your data; this allows you to manage the data from one control platform as you do any of the other footprints in the enterprise," Moser said.

Security, data growth and I/O bottlenecks are consistently the three most important ongoing problems in storage IT. This new HDS system addresses all three, with a special focus on the data growth issue.

"Most IT shops deploy what I refer to as the 'Zombie' defense (against incessant data growth)," Moser said. "The data keeps sort of lumbering at them, it's unstoppable. So they sort of throw filing cabinets and chairs in front of it to slow it down, which is to say they'll throw extra capacity in front of it. What ends up happening is that they'll have a variety of little point solutions where they've tucked and squirreled away nuggets of data all over the system.

"Then it becomes a more complex environment, and it drives operational costs up. You end up with an alphabet soup of different solutions. It's really, really painful."

What's Inside HUS

Hitachi Unified Storage provisions, tiers, migrates and protects all data, regardless of type, throughout its life cycle, Moser said, as well as adds data efficiency capabilities such as thin provisioning and automatic tiering. Built-in automation optimizes performance.

The HUS stores object data through a unique object-based file system that intelligently adds metadata for each file and enables automated tiering and migration, fast file snapshots and clones, faster replication over WAN, and fast data searches, Moser said. Additionally, Hitachi Unified Storage supports Hitachi Content Platform (HCP) for a true object store with custom metadata and provides regulatory compliance.

The content platform controller shares HUS capacity with file and block applications from the same storage pool. When all these components are combined, Moser said, this system is far more space-efficient and cost-effective than separate/siloed object store implementations.

The unified management control works across all storage models (all HDS): block (Hitachi Virtual Storage Platform), file (Hitachi NAS Platform, Hitachi Unified Storage), content (Hitachi Content Platform) and specialized appliances (Hitachi Data Ingestor).

Moser said the system can scale its capacity from small starter-type deployments of 50TB or so to nearly 3PB without affecting performance by using Hitachi Dynamic Provisioning. The system automatically corrects performance issues and provisions more quickly with its dynamic virtual controllers.

The HUS uses 256TB file systems and a single name space to reduce administrative effort.

Chris Preimesberger is eWEEK Editor of Features and Analysis and writes the Storage Station blog. Twitter: @editingwhiz

Chris Preimesberger

Chris J. Preimesberger

Chris J. Preimesberger is Editor-in-Chief of eWEEK and responsible for all the publication's coverage. In his 13 years and more than 4,000 articles at eWEEK, he has distinguished himself in reporting...