[rust] Hash-distribute log-table appends by declared bucket key#3643
Open
fresh-borzoni wants to merge 2 commits into
Open
[rust] Hash-distribute log-table appends by declared bucket key#3643fresh-borzoni wants to merge 2 commits into
fresh-borzoni wants to merge 2 commits into
Conversation
Member
Author
fresh-borzoni
force-pushed
the
fix-log-append-bucket-key-hashing
branch
2 times, most recently
from
July 14, 2026 03:17
5793b64 to
5616a40
Compare
fresh-borzoni
force-pushed
the
fix-log-append-bucket-key-hashing
branch
from
July 15, 2026 13:15
5616a40 to
3adea9e
Compare
charlesdong1991
left a comment
charlesdong1991
left a comment
Contributor
There was a problem hiding this comment.
very nice! I left two nit comments
Comment on lines
+202
to
+203
| let row = ColumnarRow::new(Arc::clone(&batch_arc), Arc::clone(&row_type), i, None)?; | ||
| let (bucket, key) = router.bucket_of(&row, num_buckets)?; |
Contributor
There was a problem hiding this comment.
Nit: it might not be very efficient when row numbers are large given this will do full typed-batch builds on each iteration, because it will also downcast each column? maybe can utilize set_row_id in ColumnarRow and move the cursor instead?
| return self.send_arrow_batch(batch, physical_table_path, None); | ||
| }; | ||
| if batch.num_rows() == 0 { | ||
| return self.send_arrow_batch(batch, physical_table_path, None); |
Contributor
There was a problem hiding this comment.
so empty batch will result in sticky assigner, but if the first batch is empty, will this mistake all following keyed appends?
maybe we can add a test to see how it behaves?
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
2 participants
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.
closes #3644
Log tables with a declared bucket key weren't hashing appends by it and every row in a flush landed in one bucket (sticky), unlike the KV/upsert path and the Java client, so bucket-key co-location was silently broken.
Now the append path hashes by the bucket key like Java:
append(row)hashes each row, andappend_arrow_batchsplits a batch across buckets (single-bucket batches send as-is, mixed batches fan out per bucket).Tables without a bucket key are unchanged.