Spark Join Strategies on Databricks: Broadcast, Sort-Merge, Shuffle Hash, and What AQE Changes

Most slow PySpark jobs I've debugged weren't slow because of bad code. They were slow because Spark...