Content-addressable storage. (Storage as I See it).

Modernly, the vast majority of all digital information is fixed content information in its final form. From MRIs and X-rays to check images, broadcast content, and completed CAD/CAM designs, fixed content is an important class of data. But some vendors assert that existing storage architectures aren't optimized for fixed content. Much of the most valuable content is stored offline, where you can't readily access it.

Thus emerged CAS--content-addressed storage. Content addressing is supposed to simplify fixed content storage and retrieval. With CAS, you get fast, affordable online access to your fixed content assets and benefit from faster application development and deployment.

EMC was first in the CAS space with its Centera architecture. EMC contrasts its approach with NAS and SAN, which it identifies as location based and optimized for frequently changing data. This, EMC says, builds complexity that is unnecessary for fixed content. EMC notes that content-based addressing found in Centera eliminates the need for applications to understand and manage the physical location of information, while introducing virtually no complexity, regardless of scale.

The software element, called CentraStar, incorporates a content-addressing intelligence that calculates a unique address based on the actual content of every stored object. This digital fingerprint serves as a permanent and portable "claim check" that applications use to retrieve objects. Retrieval requires no knowledge of the storage environment or physical location of the objects. This simplifies the task of developing Centera-integrated applications and accessing and managing huge numbers of objects, and ensures verifiable accuracy of the content for data integrity, authentication, and other purposes.

CAS is also at the core of the Axion architecture from Avamar. This newer company notes that its Axion software maximizes storage and network efficiency by finding small, logical sequences in files, objects, and databases and storing each unique logical sequence only a single time across an entire enterprise. Sounds a good deal like lossless data compression. Edited files, copied attachments, shared applications, and even daily changing databases only present a small amount of new data to an Axion system.

To identify and eliminate redundancy, Axion uses a CAS architecture. The storage address for any data element stored in the Axion system is generated by an analysis of the data itself, in a process that ensures the authenticity of the data. This unique "content address" can be used to refer to repeated data elements to eliminate redundancy, dramatically reducing Axion disk-storage requirements.

Furthermore, the content address provides direct access to data elements for immediate restore. Conventional backup solutions use a centralized index to store and locate data; this centralized index grows increasingly complex and difficult to manage as data volumes increase. CAS simplifies this challenge by using a direct-access methodology, enabling the Axion system to scale to petabytes in capacity, with no centralized index.

The time for CAS is certainly here, if for no other reason than the storage management push to simplify. The number of trained storage administrators has not increased, and IT centers need all the help they can get. All data is not created equal, and fixed content can benefit from the simplification offered by CAS.