A: Before PCs and the Internet, the ability to create, manage and store information was mostly limited to a set of management-defined applications. These transactional line-of-business systems define the content, structure, interrelationships and availability of the information they create, often referred to as structured information.
Today, more than half the world's information is created outside this environment, by applications and technologies not even dreamt of in the first decades of information storage. It is email, voicemail, x-rays, signed contracts, scanned checks, MP3s, digital photos and video clips — it's the Internet itself. Unlike transactional data, this reference information is unstructured and its content is unique and fixed.
Increasingly, the sheer bulk of the generated information, both in size and number of objects, has been overwhelming the traditional storage systems that have been the mainstay of general-purpose computing for over 30 years. These hierarchical file systems use an operator-defined tree-like classification to keep track of where information is located: which network, which server, which folder inside so many other folders. Hence they can be described as location-based.
As the increase in content storage continues, location-based storage becomes a burden. If one folder name in a hierarchy is changed, the stored content could be lost. By definition, reference information can be uniquely identified by its own content. So why use an outdated storage ideology. It would seem logical that a reference information storage solution would take this into account, but it took input from outside the conservative storage industry to make it happen. Paul Carpentier created such a solution, and along with it the multi-billion dollar category of content-addressed storage — CAS.
|
|
email us © Copyright 2008 Caringo, Inc. All rights reserved.
