In initial talks with Hortonworks, was told that I could add arbitrary data to apache atlas (something along the lines of "...applies to all hadoop API (eg. hadoop cli, spark, etc)"). Having installed HDP 3.1 for evaluation, this appears to not the case. We have ETL processes that bring in various forms of flatfile data (eg. parquet, tsv, etc.) and for none of which have I figured out how to get to show up on atlas. Is there any way to do this? Basically, would like to have this data be taggable by Ranger and discoverable via Atlas.