After merging your PR, Conbench analyzed the 5 benchmarking runs that have been run so far on merge-commit fb138c1.

Benchmarks with performance regressions

There were 0 possible performance regressions, according to the lookback z-score method.

All benchmark runs analyzed:

• Commit Run on test-mac-arm at 2024-04-16 05:57:44Z
• Commit Run on ec2-t3-xlarge-us-east-2 at 2024-04-15 22:42:20Z
• Commit Run on ec2-m5-4xlarge-us-east-2 at 2024-04-16 03:57:02Z
• Commit Run on arm64-t4g-linux-compute at 2024-04-16 02:59:48Z
• Commit Run on arm64-m6g-linux-compute at 2024-04-16 09:06:59Z

Unstable benchmarks with performance regressions

The following benchmark results indicate a possible performance regression, but are known to sometimes produce false positives when applying the lookback z-score analysis.

• Commit Run on test-mac-arm at 2024-04-16 05:57:44Z

  • ShortVectorInsert (C++) with params=<STD_VECTOR(std::shared_ptr)>, source=cpp-micro, suite=arrow-small-vector-benchmark

• Commit Run on ec2-m5-4xlarge-us-east-2 at 2024-04-16 03:57:02Z

  • setValidityBitToZeroBenchmark (Java) with source=java-micro, suite=arrow.vector.BitVectorHelperBenchmarks

• Commit Run on arm64-m6g-linux-compute at 2024-04-16 09:06:59Z

  • RecordReaderReadAndSkipRecords (C++) with params=Repetition:0/BatchSize:10/LevelsPerPage:80000, source=cpp-micro, suite=parquet-column-reader-benchmark