Generation Data Group is a group of sequential or functionally related datasets. They are handled periodically by adding a new generation, discarding the oldest generation, and sometimes retaining previous generations.
