andygrove commented on issue #330: URL: https://github.com/apache/datafusion-ballista/issues/330#issuecomment-5004177283
Closing as superseded. This checklist is from 2022 and predates several DataFusion major versions; the specific items are largely about casting bugs in the benchmark itself rather than in distributed execution, and they are no longer diagnosable in the form recorded here. Benchmark correctness is now gated automatically on every PR rather than tracked by hand: - `TPC-H SF10 (all queries, AQE off + on)` - `TPC-DS SF1 (all queries, static planner)` Both are green on `main`. Row counts are also recorded alongside timings in the Contributors Guide benchmarking page, so a fast wrong answer is visible rather than silent: https://datafusion.apache.org/ballista/contributors-guide/benchmarking.html Correctness problems found since have their own issues (for example #2046 for TPC-DS divergences under distributed execution). Please open a focused issue for any query that returns wrong results today. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
