Are the final results of a map-reduce operation indexed?

Question

I'm trying to find out how I can iterate over the final results of a map reduce operation, so I guess there must be some sort of index into the map reduce results?

Answer 1

For Hadoop MapReduce the output from the Reduce stage is a file sorted by key. To iterate over the results from a MapReduce job in Hadoop you would either need to write you own code to read these files or use another MapReduce job. It would depend what you want to do with the results.

The type of file the job will create is configurable, so it could be a plain text file or sequence files.