timothy-e commented on PR #18741: URL: https://github.com/apache/pinot/pull/18741#issuecomment-4753193539
This is exciting! I haven't looked through the code yet, but some higher-level thoughts > no hinted joins 1. It's pretty common to have a lookup join hint, but otherwise not particularly care about join order. Does 'no hinted joins' mean that we won't reorder queries that have any hinted joins? > hint veto 2. Can you elaborate more on what this means? > upsert/dedup row counts marked LOW confidence (physical docs over-count logical rows), ... The planner treats LOW/UNKNOWN confidence as "no stats". 3. Will we improve this once we have support for number of distinct values column stats? > consuming segments downgrade confidence. 4. What does downgrade mean? What confidence values do we have? 5. Also, I vote for moving the independent fixes out to a seperate PR. Maybe we'll want to revert this PR, and it would be nice to not have to revert the other fixes too. I think it'd be nice to split this up a little bit more, it's harder to find and understand the logic amidst all the plumbing. -- 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]
