We will now continue with discussing Map/Reduce and how MongoDB can leverage it. Map/Reduce is a framework which allows to parallelize the processing of large and very large datasets across many physical or virtual servers. This paradigm, firstly popularized by Google, has gotten a lot of traction these days, mostly because of the Big Data movement.
Many NoSQL solutions aim to support integration with Map/Reduce frameworks but MongoDB goes further than that and provides its own Map/Reduce implementation integrated into the MongoDB server, available for everyone to consume.