Simple example of reduce-side join in mrjob

mrjob_join.py

from mrjob.job import MRJob

class JoinExample(MRJob):
    def mapper(self, id, record):
        # Use both large files as input. If you have orders and
        # customers, you'll have as input either
        #   order_id, order_data
        # or
        #   customer_id, customer_data
        # In this case, I assume both have a customerID field to join
        # on and that you'll be able to differentiate them in the
        # reducer
        yield record['customerId'], record

    def reducer(self, customerID, records):
        for record in records:
            if is_customer_record(record):
                # do something with the customer info
            else:
                # do something with the order info
        yield customerID, new_data

if __name__ == '__main__':
    JoinExample.run()