After merging your PR, Conbench analyzed the 4 benchmarking runs that have been run so far on merge-commit 49bad4e.

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 2024-07-30 16:59:59Z
• Commit Run on ec2-c6a-4xlarge-us-east-2 at 2024-07-30 08:30:08Z
• Commit Run on arm64-t4g-linux-compute at 2024-07-30 10:38:52Z
• Commit Run on ec2-m5-4xlarge-us-east-2 at 2024-07-30 10:44:03Z

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 2024-07-30 16:59:59Z

  • TimestampParsingISO8601 (C++) with params=TimeUnit::NANO, source=cpp-micro, suite=arrow-value-parsing-benchmark
  • TimestampParsingISO8601 (C++) with params=TimeUnit::SECOND, source=cpp-micro, suite=arrow-value-parsing-benchmark
  • BM_RleDecoding (C++) with params=1024/16, source=cpp-micro, suite=parquet-column-io-benchmark
  • TimestampParsingISO8601 (C++) with params=TimeUnit::MICRO, source=cpp-micro, suite=arrow-value-parsing-benchmark
  • BM_WriteInt64Column (C++) with params=Repetition::REPEATED/1048576, source=cpp-micro, suite=parquet-column-io-benchmark
  • TakeStringRandomIndicesWithNulls (C++) with params=4194304/0, source=cpp-micro, suite=arrow-compute-vector-selection-benchmark
  • AsciiUpper (C++) with source=cpp-micro, suite=arrow-compute-scalar-string-benchmark
  • TableSortIndicesInt64Wide (C++) with params=1048576/4/2/32, source=cpp-micro, suite=arrow-compute-vector-sort-benchmark
  • ListSliceInt64ListWithStop (C++) with params=65536/10, source=cpp-micro, suite=arrow-compute-scalar-list-benchmark

• Commit Run on ec2-c6a-4xlarge-us-east-2 at 2024-07-30 08:30:08Z

  • BM_RleEncoding (C++) with params=4096/8, source=cpp-micro, suite=parquet-column-io-benchmark
  • BM_RleEncoding (C++) with params=4096/16, source=cpp-micro, suite=parquet-column-io-benchmark
  • BM_RleEncoding (C++) with params=65536/16, source=cpp-micro, suite=parquet-column-io-benchmark
  • BM_RleEncoding (C++) with params=32768/16, source=cpp-micro, suite=parquet-column-io-benchmark
  • BM_RleEncoding (C++) with params=32768/8, source=cpp-micro, suite=parquet-column-io-benchmark
  • BM_RleEncoding (C++) with params=65536/8, source=cpp-micro, suite=parquet-column-io-benchmark
  • BM_RleEncoding (C++) with params=1024/8, source=cpp-micro, suite=parquet-column-io-benchmark
  • BM_RleEncoding (C++) with params=1024/16, source=cpp-micro, suite=parquet-column-io-benchmark