Feature Story
|
By Leslie T. O'Neill February 26, 2008 - Massive volumes of photographs and videos, scanned document images, and other files are being created today, and they must be archived in a way that protects them and makes them easily searched for and accessible to large groups of users—forever. It is risky to use proprietary, closed source archive technology for this type of fixed-content archive because the data may become marooned on an archive "island" if the vendor leaves the business, the archiving organization loses key technical personnel, the product is significantly altered, or newer technology supersedes it. On the other hand, open source software can help prevent archived fixed content from being stranded in this way because its source code is always publicly available. It is even more attractive when you consider that open source code is easy on the budget and can reduce technical complexity. All of these advantages are increasingly important to all organizations, especially universities and government agencies, that have massive amounts of content that they must archive, tag with metadata, preserve for the ages, and keep available to millions of users. To answer this need, Sun announced today at its annual Worldwide Education and Research Conference (WWERC) in San Francisco that it has made the fixed-content object storage software of the StorageTek 5800 system--also known as Honeycomb--free and open. "The people who are responsible for preserving national treasures and making them accessible from the Web, such as the Library of Congress, are critically interested in building archives that last forever, which is challenging since they will not have their current hardware or their current archive implementation for 100 years," says Graham Lovell, Sun's senior director of storage servers and appliances. "But with this announcement, they'll gain confidence that the software code will be available and will be advanced by the community, and that they will be able to build on it themselves." As a part of its open approach to archiving, Sun is donating the source code for the Sun StorageTek 5800 system to three core communities: the OpenSolaris community for ISVs, system administrators, and storage software, the Java.Net community for Java developers and the Storage Networking Industry Association (SNIA) for use with standards developments around the eXtensible Access Method (XAM) object access standard. In addition, the popular Fedora Commons open source group plans to add its software to these communities as well as use the Honeycomb source code in product development. The Sun StorageTek 5800 is based on the Sun Solaris operating system, and is the first integrated digital archive system built with open source software. Developers can immediately download the binary code, also called Project Honeycomb, from Sun.com and run a complete implementation on their own server or even a laptop. The source code will be available on March 4, 2008. "Typically, people have been for several years ingesting fixed content--data, such as photographs or videos, that will never change--into a storage system like the StorageTek 5800," says Lovell. "They may have as many as 10 billion objects in an archive, and they don't want to redo their archives with a new vendor's system. But if customers can control how data is stored and retrieved with open source software, they don't need to worry about vendor support." A Free and Open ArchiveIn addition to eliminating the risk of vendor lock-in, Project Honeycomb makes object storage extensible and affordable. According to Lovell, customers could easily spend $50,000 on a propriety combination of hardware and software yet have no guarantee that vendor support would span the long lifetime of their archives. The StorageTek 5800 system is the only integrated storage system built on open source software that is capable of tagging and retrieving files with metadata, which makes it significantly faster to find specific content; this is a critical feature when your organization's users troll through hundreds of millions of files. A fixed-content object can be archived with rich metadata; for example, searchable tags of each file might include a description of a photograph or video, its resolution, the date the original was created, and more. Because Sun has open-sourced the core engine of the StorageTek 5800, you can use it to store and retrieve fixed objects as well as tag and search them. "There is incredible value in object storage," says Lovell, "and Project Honeycomb gives free access to object storage to anybody." Sun will continue to sell and service fully integrated solutions built on the open source code. With the purchase of the StorageTek 5800 system, you also get enhanced reliability, availability, and serviceability, including protection against data corruption and loss through advanced data integrity functionality. Ongoing Open SourcingNext, Sun will open-source the StorageTek Storage Archive Manager (SAM) and the new Key Management System (KMS) 2.0 encryption management appliance. KMS 2.0 can support as many as 30 times more devices than competing solutions, and it is the first product to offer easy, centralized key management for tape drives from multiple vendors. Continuing its leadership in contributing free and open software to the IT community, Sun is the first vendor to commit to open-sourcing a key-management system. Get StartedYou can start building an object store for your organization's library today. Here's how:
Leslie T. O'Neill writes about Sun technology and was the Test Center Managing Editor and Special Projects Editor at InfoWorld magazine. |
Podcast: Innovating@Sun
Hal Stern interviews Josh Dobies about open-sourcing the StorageTek 5800 Honeycomb System.
Listen Now
 | ||||||||||||||
 |
| ||||||||||||||