The data fabric concept has persisted over the past couple years as a means of providing centralized data access in an increasingly decentralized data landscape. Organizations have data scattered across various cloud, on-premise settings, and hybrids of the two. A data fabric, so the story goes, unites this data while granting singular access to it.
But what exactly is a data fabric, and how does it work?
The answer is twofold. According to Stardog CEO Kendall Clark, a “data fabric, at least as we envision it, gives you a data management strategy that’s an alternative to the moving and copying of data.” As such, it’s a uniform means of connecting all data and managing it as though it were in the same place.
As for implementation, one of the most effective means of achieving this advantage is via “query federation, data virtualization, and a graph data model,” Clark stated. With this approach, data is connected through a virtualization layer that supports federated queries across sources with the inherent flexibility of semantic graph schema.
Constructing a data fabric this way delivers numerous advantages for analytics, including reduced costs, greater flexibility, and improved speed, while getting better results from the data in which organizations have invested.
Read the rest of the article on Analytics Week here.