amol- edited a comment on pull request #11284: URL: https://github.com/apache/arrow/pull/11284#issuecomment-933546115
> I'm ok with the change but it is completely unrelated to file reading performance (which doesn't go through the Python conversion path). Sorry, left this as a draft branch because I couldn't see locally how the original PR changes impacted the benchmarks, so I opened a PR to tweak them and see if it had any impact on the conbench. I also tried doing a `git bisect` locally to spot the commit that introduced slowdowns, but couldn't see any real difference in the benchmarks. We will have to dig further if we want to actually find the culprit of the slowdown, I'm having an hard time proving it's related to the changes in https://github.com/apache/arrow/pull/11076 that seemed the original culprit of the slower timings according to the conbench report. -- 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]
