feat: Support RIGHT/FULL joins in NLJ memory-limited execution#21833
Open
viirya wants to merge 1 commit intoapache:mainfrom
Open
feat: Support RIGHT/FULL joins in NLJ memory-limited execution#21833viirya wants to merge 1 commit intoapache:mainfrom
viirya wants to merge 1 commit intoapache:mainfrom
Conversation
github-actions
Bot
added
sqllogictest
viirya
force-pushed
the
nlj-right-full-join-spill
branch
3 times, most recently
from
75e43cf to
eb80057
Compare
Previously RIGHT/FULL/RIGHT SEMI/RIGHT ANTI/RIGHT MARK joins were excluded from the memory-limited (multi-pass) fallback path because they need to track which right rows have been matched across all left chunks. They would OOM instead of spilling. Now all join types support the fallback: - A global right bitmap (Vec<BooleanBuffer>, indexed by right batch sequence number) accumulates matches across all left chunk passes. ReplayableStreamSource guarantees consistent batch boundaries across passes, so batch sequence numbers are stable. - In memory-limited mode, EmitRightUnmatched merges the current batch's bitmap into the global accumulator (bitwise OR) instead of emitting unmatched rows immediately. - After the last left chunk, a new state EmitGlobalRightUnmatched replays the right side one more time and uses the accumulated bitmap to emit unmatched right rows correctly. Single-pass behavior is unchanged: the global bitmap path is only active when is_memory_limited() is true. Co-authored-by: Claude Code
viirya
force-pushed
the
nlj-right-full-join-spill
branch
from
eb80057 to
57c8cb8
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Which issue does this PR close?
Rationale for this change
What changes are included in this PR?
Previously RIGHT/FULL/RIGHT SEMI/RIGHT ANTI/RIGHT MARK joins were excluded from the memory-limited (multi-pass) fallback path because they need to track which right rows have been matched across all left chunks. They would OOM instead of spilling.
Now all join types support the fallback:
A global right bitmap (Vec, indexed by right batch sequence number) accumulates matches across all left chunk passes. ReplayableStreamSource guarantees consistent batch boundaries across passes, so batch sequence numbers are stable.
In memory-limited mode, EmitRightUnmatched merges the current batch's bitmap into the global accumulator (bitwise OR) instead of emitting unmatched rows immediately.
After the last left chunk, a new state EmitGlobalRightUnmatched replays the right side one more time and uses the accumulated bitmap to emit unmatched right rows correctly.
Single-pass behavior is unchanged: the global bitmap path is only active when is_memory_limited() is true.
Co-authored-by: Claude Code
Are these changes tested?
Unit tests and e2e tests
Are there any user-facing changes?
No