falcon

Hi,

First of all, thank you all for Falcon. It's awesome.

However, it looks like the Apache Drill connector only works with S3 buckets. Nonetheless, Drill itself is able to query from lots of other places (such as a local filesystem, HDFS, MongoDB...).

Adding generic support for Apache Drill (without the need for S3 credentials) would be great. It would expand Falcon capabilities a lot

Hey! When I was reading the docs and I saw api.background.task, the first thing that popped into my head was "but you can just use asyncio.create_task or loop,.run_in_executor for that without inventing something new!". But then I noticed that it also passes the context vars to the synchronous backround stuff, which is nice. Wonder why that's not the case in the stdlib, cause it would be nic

Description

The top-level README does not provide any information or references for running the stack. That information was moved out into a separate docs file in #427. It might be beneficial for first-time contributors to reference that document from the top-level README, since they may not know where to look for it.

Additional context