Needs more docs

[drcrallen]

Before any public release, this needs more documentation.

[l15k4]

Hey, could you please include some Rationale of this extension? Like what is the reason of its existence, what problems does it solve, etc.

[xvrl]

@l15k4 we moved all our internal batch processing from Hadoop to Spark, so we also wanted Druid indexing run on Spark to avoid maintaining a separate Hadoop cluster just for Druid