spark
spark copied to clipboard
[SPARK-40487][SQL] Make defaultJoin in BroadcastNestedLoopJoinExec running in parallel
What changes were proposed in this pull request?
Currently, the defaultJoin method in BroadcastNestedLoopJoinExec collects notMatchedBroadcastRows firstly, then collects matchedStreamRows. The two steps could run in parallel instead of serial.
Why are the changes needed?
Make defaultJoin in BroadcastNestedLoopJoinExec running in parallel
Does this PR introduce any user-facing change?
NO
How was this patch tested?
UT
Can one of the admins verify this patch?
@cloud-fan @c21
makes sense, also cc @maryannxue @somani
Merged to master.