hadoop - Is key necessary in MapReduce? -


in case not need reduce function. assuming map function should not worry choosing , splitting input text file key-value pair.

yes does. mappers output key,value pair. if don't want use reducer, can write map output directly file system, or can use identity reducer. if you're not interested in key is, can maybe assign default key. if can share more details you're trying do, can out better.


Comments