[Fiahil]

Databricks has an extremely bad API. So, sure, your Spark jobs might be a little bit faster some times, but why would you use it if you can't even read logs of running jobs?