Aman Sinha has posted comments on this change. ( http://gerrit.cloudera.org:8080/20009 )
Change subject: IMPALA-12183: Fix cardinality clamping across aggregation phases ...................................................................... Patch Set 5: Code-Review+1 (1 comment) LGTM. Will bump the vote once the tests are completed and also give other reviewers some time to review. http://gerrit.cloudera.org:8080/#/c/20009/4/testdata/workloads/functional-planner/queries/PlannerTest/tpcds-processing-cost.test File testdata/workloads/functional-planner/queries/PlannerTest/tpcds-processing-cost.test: http://gerrit.cloudera.org:8080/#/c/20009/4/testdata/workloads/functional-planner/queries/PlannerTest/tpcds-processing-cost.test@322 PS4, Line 322: 3373800 > I put some debugging logs and these are the num group estimates and the cap Sounds good. Just expanding a bit on the overall processing cost of this operator so we have a common understanding of this: - The UNION below the AGGREGATE #127 produces 276K rows. - AGGREGATE #127 has 5 agg classes - each class will internally process all 276K rows but due to the group-by each class produces <= 276K rows and the total output cardinality of AGGREGATE #127 is 562K rows. This corresponds to the processing cost of 8308890. i.e processing 276K * 5 = 1380K rows costs 8308890 - AGGREGATE #215 is consuming the already grouped output per-class. Class 0 consumes Class 0 coming from #127 etc. so applying the per class cardinality you get total cost 3373800 which is 40% of the cost of #127. -- To view, visit http://gerrit.cloudera.org:8080/20009 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I1d414fe56b027f887c7f901d8a6799a388b16b95 Gerrit-Change-Number: 20009 Gerrit-PatchSet: 5 Gerrit-Owner: Riza Suminto <[email protected]> Gerrit-Reviewer: Abhishek Rawat <[email protected]> Gerrit-Reviewer: Aman Sinha <[email protected]> Gerrit-Reviewer: Impala Public Jenkins <[email protected]> Gerrit-Reviewer: Kurt Deschler <[email protected]> Gerrit-Reviewer: Riza Suminto <[email protected]> Gerrit-Comment-Date: Wed, 07 Jun 2023 18:03:15 +0000 Gerrit-HasComments: Yes
