jt-kloudfuse commented on pull request #8228: URL: https://github.com/apache/pinot/pull/8228#issuecomment-1045929106
> Since we are only optimizing the case of all order-by columns are sorted, we can have the check within `getNextBlock()` and add a new method `computeAllSorted()` to only add the first `limit` records into the priority queue. Mixing the logic with existing computation will introduce extra overhead, and hard to read I did consider doing that. At the moment, getNexBlock has a conditional split for all/partially ordered columns (where computeAllOrdered is where all columns in select are also in the order by if I understand correctly ) If we introduce a case for when all orderBy columns are pre-sorted (early termination case) then wouldn't we again have a allOrdered/partialOrdered case within that as well? -- 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]
