Evaluate and possibly integrate an IPython notebook environment in the single machine package

[bashir2]

Currently, our single machine deployment package does not include any notebook or SQL environment. Instead we provide the Thrift server and clients can connect to it through JDBC. Once we start to rely more on FHIR views for query needs, having a python/notebook environment becomes even more important. We should evaluate different options and possibly integrate one. There are some options here.

[bashir2]

To clarify: The image we add here should include required JDBC drivers (and Java deps), FHIR-views as a packaged library (with Spark support), and possibly other useful libraries that do not come with the upstream image. That said, we should try to limit the overhead for maintaining this image.