alamb commented on PR #21480: URL: https://github.com/apache/datafusion/pull/21480#issuecomment-4229247865
> This removes a redundant step, and the change looks reasonable. However, I’m curious how this step ends up consuming significant CPU time — building a predicate per file doesn’t seem very expensive. Do you have a micro-benchmark that demonstrates the impact on a representative workload? @2010YOUY01 roughly 10-20% of the flamegraph in the PR description is building the page pruning predicate <img width="1920" height="840" alt="575753025-ae7797eb-2b44-4b2f-bf41-28e4e99f2386" src="https://github.com/user-attachments/assets/541acc5c-8e0f-42f9-9f6d-f21b9fbd4adb"; /> -- 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]
