What is a broadcast join used for?
- A Sending a small table to every executor so a large table can be joined without shuffling
- B Joining two very large tables
- C Streaming joins only
- D Joining across clusters
Answer
Sending a small table to every executor so a large table can be joined without shuffling
It eliminates the shuffle entirely for the large side, which is one of the most effective Spark optimisations available.





