Cross join

Cross join matches every row from the left with every row from the right, generating a Cartesian cross product. This is to be used with caution, as it is the worst performing join, to be used in specific use cases only:

This will output all temperatures for all cities, generating 6 x 6 records (36 output records). This join is usually not used, as the output can be very large, and it is not that useful in most cases.

Thus, we can implement different joins using the multiple mapper approach. 

The multiple mappers reducer job seen earlier is a good example of a join pattern.

Depending on the use case, a join pattern can be customized to generate the intended output.