Important issues:
- use jdk 1.8
- sbt archetype (like holdens) is awkward, for students you can make your own
- sbt is slower than maven+scala
How to run:
- download Spark (2.3.4, for hadoop 2.7)
- add Spark path to bin
- put the payload (json winemag)
spark-submit --master local[*] --class json_reader_matkheev.sparkProject.JsonReader scala-2.11/sparkProject-assembly-0.0.1.jar winemag-data-130k-v2.json