Develop Genny workload that populates a new replica set for testing.

Workload should create a set of documents with a predetermined size distribution. The distribution should use a mix of small and larger document sizes.  A zipfian distribution with lots of small documents and a smaller number of big ones would probably be sufficient. 

When this task is complete, we should have a (basic) Genny workload that sets up a PSA replica set for testing and populates it with a set of files using the workload developed here.