Map-Reduce

Is a technique that has two tasks:

1. Map tasks (Splits & Mapping)
2. Reduce tasks (Shuffling, reducing)

For example given an input where we need to get the frequency count of each word:

• First, the mapping will map the frequency of each word in the input splits.
• Shuffling will then sort the word so that it's easier for the reducer to aggregate the data.
• Reducer simply aggregate all the data to the final output

Hadoop is an open source framework for map-reduce