Repository navigation
Conversation
… testing Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…pache/arrow-rs#11070) Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Contributor
Author
|
run benchmark clickbench_partitioned |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing alamb/test_arrow_main_oct15 (4fd151c) to b645de0 (merge-base) diff Run configurationrun benchmark clickbench_partitionedResults will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing alamb/test_arrow_main_oct15 (4fd151c) to b645de0 (merge-base) diff Run configurationrun benchmark clickbench_partitionedCPU Details (lscpu)Details
Resource Usageclickbench_partitioned — base (merge-base)
clickbench_partitioned — branch
File an issue against this benchmark runner |
* `interleave` no longer repeats a shared data buffer for byte view arrays (apache/arrow-rs#11281), so the spill `compact_view_array` tests build the repeated-buffer array by hand * Casting a naive timestamp to a named time zone now resolves DST-ambiguous and nonexistent local times like PostgreSQL instead of erroring (apache/arrow-rs#11038): update `timestamps_timezone.slt` Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
|
Thank you for opening this pull request! Reviewer note: cargo-semver-checks reported the current version number is not SemVer-compatible with the changes in this pull request (compared against the base branch). Details |
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #26384 +/- ##
==========================================
+ Coverage 82.85% 82.88% +0.03%
==========================================
Files 1151 1153 +2
Lines 456012 457501 +1489
Branches 456012 457501 +1489
==========================================
+ Hits 377824 379218 +1394
- Misses 55235 55271 +36
- Partials 22953 23012 +59 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
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
3 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.
Which issue does this PR close?
mainNote: this is a testing PR and is NOT intended to be merged. It pins DataFusion to the current head of arrow-rs
mainto find any problems before the next arrow release.Rationale for this change
The upcoming arrow/parquet release contains 250+ commits since
60.0.0, including several large parquet reader changes (ParquetPushDecoder::scan_plan(apache/arrow-rs#10555), fused same-projection predicates (apache/arrow-rs#10859), releasing row group bytes when the push decoder is done with them (apache/arrow-rs#11235), partial page index population (apache/arrow-rs#11157), andRowSelection/ filter performance work). Running the DataFusion CI againstmainahead of the release lets us find regressions while they are still cheap to fix upstream.What changes are included in this PR?
Add a
[patch.crates-io]section pinning all arrow/parquet crates to apache/arrow-rs@dd24011 (mainas of 2026-10-15)Cargo.lockupdateReplace the one use of
FixedSizeListArray::value_offset(inunnest.rs) withvalue_offset_at: the former was deprecated in fix(arrow-array): deprecate FixedSizeListArray::value_offset, which wraps past i32::MAX arrow-rs#11070 and failscargo clippy -D warningsThat is the only code change needed: DataFusion otherwise compiles unchanged against arrow-rs
main.Findings so far
CI against
mainfound two intentional behavior changes in arrow-rs that DataFusion tests had pinned the old behavior of. Both are updated in this PR (second commit) and will be needed in the real upgrade PR:interleavereuses shared buffers for byte view arrays (Reuse buffers when interleaving byte views arrow-rs#11281). Threespill::tests::test_gc_*tests built an array with the same data buffer listed once per input viainterleave, which now dedups the buffers itself, so there was nothing left forcompact_view_arrayto do. The test helper now assembles the repeated-buffer array by hand (the situation still arises from other kernels and older batches).Timestamp(_, None)to a named timezone errors on DST boundaries #25084 for the cast /AT TIME ZONEpaths).timestamps_timezone.sltexpectedCannot cast timezone to different timezoneerrors; it now expects the PostgreSQL answers. The string parsing paths (arrow_cast('2024-03-10T02:30:00', 'Timestamp(Nanosecond, Some("America/Denver"))'),::timestamptzunder a session zone) still fail witherror computing timezone offset.Benchmarks:
clickbench_partitionedon this PR shows no change againstmain(see the bot comment below).What is the testing strategy for this PR?
CI runs against the pinned arrow-rs
main.Are there any user-facing changes?
No