zhuqi-lucas commented on PR #23369: URL: https://github.com/apache/datafusion/pull/23369#issuecomment-4904983920
CI ClickBench numbers on this PR — cc @alamb @adriangb. Change: flip `pushdown_filters` default → `true`, plus a small gate in `ParquetSource::try_pushdown_filters` that declines when the projection has < 3 non-filter columns (falls back to FilterExec). | | HEAD | This PR | |---|---|---| | Faster | – | 5 | | Slower | – | 4 (likely noise) | | Total | 19549 ms | 16735 ms (14% faster) | | Q23 | 3107 ms | 147 ms (21.12×) | Basically no regression — the 4 "slower" queries all have wide per-iteration variance in this run. That said, I agree the adaptive stack ([#22144](https://github.com/apache/datafusion/pull/22144)) is the right long-term solution — this PR is only a stopgap. -- 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]
