MapReduce is a Google-implemented framework for calculating very large amounts of data from multiple petabytes in a short period of time. MapReduce is used by Google to indicate large amounts of documents for its search engine, and analysis of visitor numbers and visitor behavior of websites (Google Analytics).
MapReduce can process a lot of data in a short time by splitting a large task into shared tasks. Those part-time tasks are distributed across multiple computers. The computers then carry out the sharing task simultaneously (distribution). This is done by using the functions folder and fold, known from the functional programming languages. If one of the computers passes while performing its sub task, that part job is automatically taken over by another computer. This makes the system very robust.
Hadoop is a MapReduce-inspired open source Java framework for building data-intensive distributed applications. Externe link
wiki
