Tuesday, April 26th, 2011
Hadoop is a new paradigm for data processing that scales near linearly to petabytes of data. Commodity hardware running open source software provides unprecedented cost effectiveness. It is affordable to save large, raw datasets, unfiltered, in Hadoop’s file system. Together with Hadoop’s computational power, this facilitates operations such as ad hoc analysis and retroactive schema changes. An extensive open source tool-set is being built around these capabilities, making it easy to integrate Hadoop into many new application areas.