DATABRICKS-SPARK Question 19

Community Discussion

Written by other learners. Comments can disagree with the marked answer — check the official documentation when in doubt.

• Chose A0

  - F.broadcast wraps a DataFrame to hint that a broadcast hash join should be used — Spark sends the small DataFrame to every executor, eliminating shuffle of the large side. Default broadcast threshold is spark.sql.autoBroadcastJoinThreshold (10 MB). Caching and repartition don't force broadcast.