austin3dickey commented on issue #38438: URL: https://github.com/apache/arrow/issues/38438#issuecomment-1777684208
I was able to reproduce this on my M1 Mac. Steps: - build Arrow and pyarrow using [this](https://github.com/voltrondata-labs/arrow-benchmarks-ci/blob/89b79fbbe845d4d1e3c8b481f540de5cf7d14af9/buildkite/benchmark/utils.sh#L122) as reference (python 3.8) - Instead of running `python -m buildkite.benchmark.run_benchmark_groups`, I manually ran the `dataset-serialize` benchmark by: - cloning https://github.com/voltrondata-labs/benchmarks - running `pip install -e .` in that repo - creating a data dir and saving it to `BENCHMARKS_DATA_DIR` - creating a temporary output dir and saving it to `BENCHMARK_OUTPUT_DIR` (only required since I'm using Darwin) - setting `export DRY_RUN=true` so it doesn't try to post to Conbench - running `conbench dataset-serialize ALL --iterations=6 --all=true --drop-caches=true` with sudo permissions This was able to run one case, and generate a segfault on the second case. Logs look the same as [here](https://github.com/voltrondata-labs/arrow-benchmarks-ci/issues/166#issuecomment-1775365756). -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
