What is the difference between a zero reducer and identity reducer in Hadoop Mapreduce

Question

I'm a newbie to Hadoop and Big Data Ecosystem. I want to know the difference between a zero and an identity reducer. Help needed.

Answer 1

A Zero reducer as the name suggests omits the reducing step and the final output is only the map part of MapReduce. It is also known as a Map-only step.

As far as  Identity Reducer is concerned, the input Key, Value pairs are just dumped into the output as it is without any aggregation, except the data is sorted based on the key. Here, shuffle & sort happens except the aggregation. Hence, Identity Reducer is used if you only want to sort the input data from map output.