Aman Sinha has posted comments on this change. ( http://gerrit.cloudera.org:8080/22047 )
Change subject: IMPALA-2945: Account for duplicate keys on multiple nodes preAgg ...................................................................... Patch Set 15: Code-Review+2 (1 comment) A comment below for future consideration. Overall, LGTM. On the other patch which estimates the cardinality using the formula in IMPALA-2945, I just want to give it some more thought. Will get back on that patch. http://gerrit.cloudera.org:8080/#/c/22047/15/testdata/workloads/functional-planner/queries/PlannerTest/processing-cost-plan-admission-slots.test File testdata/workloads/functional-planner/queries/PlannerTest/processing-cost-plan-admission-slots.test: http://gerrit.cloudera.org:8080/#/c/22047/15/testdata/workloads/functional-planner/queries/PlannerTest/processing-cost-plan-admission-slots.test@102 PS15, Line 102: 2.30K Ideally, the right value for this should have been closer to NDV * num_instances = 6 * 3.= 18. But I checked with Riza offline and he clarified that the number of estimated instances is 384 (see line 95 above), so 6 * 384 = 2304. Given that calculation, I am good with it but just leaving a comment here for future reference that the reason for the significant cardinality overestimation is actually the num instances. It would be good to have a separate Jira for that for future work. -- To view, visit http://gerrit.cloudera.org:8080/22047 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I04c563e59421928875b340cb91654b9d4bc80b55 Gerrit-Change-Number: 22047 Gerrit-PatchSet: 15 Gerrit-Owner: Riza Suminto <[email protected]> Gerrit-Reviewer: Aman Sinha <[email protected]> Gerrit-Reviewer: Csaba Ringhofer <[email protected]> Gerrit-Reviewer: David Rorke <[email protected]> Gerrit-Reviewer: Impala Public Jenkins <[email protected]> Gerrit-Reviewer: Michael Smith <[email protected]> Gerrit-Reviewer: Riza Suminto <[email protected]> Gerrit-Comment-Date: Thu, 16 Jan 2025 03:07:21 +0000 Gerrit-HasComments: Yes
