After merging your PR, Conbench analyzed the 3 benchmarking runs that have been run so far on merge-commit 7540ad8.

Benchmarks with errors

These are errors that were caught while running the benchmarks. You can click each link to go to the Conbench entry for that benchmark, which might have more information about what the error was.

• Commit Run on arm64-t4g-2xlarge-linux at 2026-01-28 11:35:15Z
  • dataset-read (Python) with async=True, dataset=nyctaxi_multi_parquet_s3, pre_buffer=true

Benchmarks with performance regressions

There weren't enough matching historic runs in Conbench to make a call on whether there were regressions or not.

To use the lookback z-score method of determining regressions, there need to be at least two historic runs on the default branch which, when compared to one of the runs on the contender commit, are on the same hardware, and have at least one of the same benchmark case and context pairs.

All benchmark runs analyzed:

• Commit Run on arm64-t4g-2xlarge-linux at 2026-01-28 11:35:15Z
• Commit Run on amd64-c6a-4xlarge-linux at 2026-01-28 09:38:50Z
• Commit Run on amd64-m5-4xlarge-linux at 2026-01-28 11:33:54Z