mudit-97 commented on PR #9479: URL: https://github.com/apache/iceberg/pull/9479#issuecomment-1903196974
> I spent little time and understood this PR. > > Basically you want to enable record level filtering (and you have observed the benefits with this POC PR) for vector reader instead of high level row-group filtering. > > You want to know why it is not enabled for vector reader but enabled for plain ParquetReadBuilder without reader functions (readerFunc and batchedReaderFunc). > > I don't have answer for this history. I can check in iceberg slack or mailing list. We might have to run all benchmarks and understand whether any side effects of enabling record level filtering. Yes @ajantha-bhat , we were running TPCDS benchmarks there we saw in some places, this was causing some degradation while in some places, it was improving the results. That's why we wanted to understand once whether we are going in right direction with this or not -- 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: issues-unsubscr...@iceberg.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org For additional commands, e-mail: issues-h...@iceberg.apache.org
