panbingkun commented on PR #48804: URL: https://github.com/apache/spark/pull/48804#issuecomment-2469429947
> No, it's opposite. If that's true, we definitely should proceed it independently in order to measure how much they are affecting, @stevomitric . Please make a benchmark update PR first. > > > The regression was partially due to benchmarks as well. The CollationFactory.collationNameToId function should only be called fixed amount of times per query, which shouldn't measurably impact the performance of the query. > > So i think this is the right PR for it, since it was the leading contributor to the numbers seen in the results. Yes, I also agree, because in the previous PR (https://github.com/apache/spark/pull/48222) that may have generated performance regression, we did not see any modifications to `CollationBenchmark`. Therefore, in order to address this issue more clearly, can we do it separately? Thanks! -- 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]
