Article

Sun debuts its Honeycomb data archive

Beth Pariseau
After years of preannouncements and more than a year of alpha and beta testing, Sun Microsystems Inc.'s Sun/StorageTek 5800 data archiving system, better known as Honeycomb, is finally generally available.

The system will have to compete with data archiving products that established themselves during Honeycomb's long incubation, from fixed-content systems such EMC Corp.'s Centera, Hewlett-Packard Co.'s Integrated Archive Platform (IAP, formerly RISS) and Hitachi Data Systems' Hitachi Content Archiving Platform (HCAP), to targeted email and file archiving applications.

Sun said Honeycomb will differentiate itself because it uses a

    Requires Free Membership to View

RAIN architecture based on "cells" of eight or 16 1U commodity servers running open source Solaris. It also runs Reed-Solomon hashing algorithm for fixed-content archiving, which assigns a unique address to each data object for compliance and search purposes.

More on data compliance and archiving
Archiving vendors beef up integration

Fixed-content data archiving products get a boost

Vanderbilt hospital fights archive growth with grid

Google buy shakes up email archiving
As promised early in its development, the 5800 system can also run applications, and Sun has opened some code to developers to create new applications through the Storage Beans program (named after Java Beans). Sun senior manager of strategy and planning Patrick Auth said these applications could include everything from image format conversion to data deduplication.

For the initial release, Storage Beans has come up with a search application that can run across all the nodes of Honeycomb, drawing on the processors in each of its servers, rather than requiring processing at the client/server level.

"We can preserve the metadata along with objects in the archive, and you won't have to pull them out of the archive to manage and manipulate the data," Auth said. Sun's stance is that during decades-long, long-term data preservation, relying on proprietary applications outside the archive to access data is a losing proposition.

Sun also claims its hardware migrations will be easier than those of its competitors because it uses industry- standard hardware. But Auth admits that Sun's software will still have to be updated to accommodate new hardware systems.

Sun left out one relatively mundane, but crucial feature: support for multisite replication. And like most Sun storage products released recently, Honeycomb uses ZFS, currently the subject of patent-infringement lawsuits with Network Appliance Inc. (NetApp). Honeycomb uses the filesystem to boot nodes and control metadata.

Storage systems product group director John Considine said Honeycomb could survive the worst-case scenario of ZFS being taken off the market. "I can't comment on the lawsuit directly," he said. "But if users are concerned, we could fall back just to the pure architecture and aren't necessarily dependent on the underlying file system for the core value proposition."

It's unclear how many users there are with the system in production to date, or how many of them are paying customers. Sun claims several user references with this release, including Stanford University, Johns Hopkins University, Oxford University, Alberta Libraries and Southampton University, but would only confirm that Stanford, Johns Hopkins and Southampton hadn't had the system donated by Sun. Contacts at those institutions were not available for interviews by press time, although a representative from Stanford (where Sun was started) spoke with SearchStorage.com at the time the university joined the early adopter program for Honeycomb last year.

Analysts said it's a good sign that Sun is executing on its plan to base as much of its products as possible on Sun IP, and its goals of releasing hybrid storage server products when it combined its storage and server product groups last month. Sun's ability to port applications and capture metadata within the system is also winning fans.

"In markets like oil and gas and pharmaceuticals, they have very large data sets that are expensive to glean, and sometimes how they were gleaned is also important," Clipper Group analyst Anne MacFarland said. "Being able to run all that locally could cut down on the costs, and being able to use parallel processing to reuse data could improve the ability to retrieve the data quickly."

According to John Webster, principal IT advisor with Illuminata Inc., the server-grid architecture in the 5800 could also be a differentiator. "There really aren't too many products out there yet, especially in service to archival storage, based on parallel processing," he said.

However, Webster said, long-term object presentation is still being developed in open source communities. "They're still figuring out ways to migrate between hardware platforms, and it's harder to do that from a software perspective with object-based storage," he said. "It's more or less still an unsolved problem at this point."

The Honeycomb system is built in cells and is available in configurations of eight servers (known as a half cell) or 16 servers (a full cell). A full cell has a capacity of 32 TB and lists for $245,000; a half-cell holds 16 TB and is priced at roughly half the cost of the full cell.


There are Comments. Add yours.

 
TIP: Want to include a code block in your comment? Use <pre> or <code> tags around the desired text. Ex: <code>insert code</code>

REGISTER or login:

Forgot Password?
By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy
Sort by: OldestNewest

Forgot Password?

No problem! Submit your e-mail address below. We'll send you an email containing your password.

Your password has been sent to: