That’s exactly what MapReduce does. The input is processed by the mapper function, and its output will be automatically sent into the reducer function. Between mappers and reducers we have the automatic shuffle phase which sends records with identical keys into one reducer call.
If you want to simply not modify the data through the map phase, just don’t specify the mapper class in your driver code. The identity function will be used for the map phase in this case. Kai Am 19.06.2014 um 11:00 schrieb Shrivastava, Himnshu (GE Global Research, Non-GE) <himnshu.shrivast...@ge.com>: > Can we take input in a mapper function and in some way pass it to reducer ? > > Regards, Kai Voigt Am Germaniahafen 1 k...@123.org 24143 Kiel +49 160 96683050 Germany @KaiVoigt