Riza Suminto has posted comments on this change. ( http://gerrit.cloudera.org:8080/21955 )
Change subject: IMPALA-13465: Trace TupleId further to reduce Agg cardinality ...................................................................... Patch Set 15: (4 comments) http://gerrit.cloudera.org:8080/#/c/21955/12/fe/src/main/java/org/apache/impala/planner/AggregationNode.java File fe/src/main/java/org/apache/impala/planner/AggregationNode.java: http://gerrit.cloudera.org:8080/#/c/21955/12/fe/src/main/java/org/apache/impala/planner/AggregationNode.java@286 PS12, Line 286: + "preaggNumGroup={} aggInfo={}", > This makes the code more fragile, and that concerns me. It's uglier, but I' Reverted to use getDisplayLabel() inside isTraceEnabled check. http://gerrit.cloudera.org:8080/#/c/21955/12/fe/src/main/java/org/apache/impala/planner/AggregationNode.java@439 PS12, Line 439: for (TupleId id : producerNode.getTupleIds()) { > Not yet. I can file one, but it probably will take sometime until we can ha Mentioned HIVE-26221 as example. http://gerrit.cloudera.org:8080/#/c/21955/13/fe/src/main/java/org/apache/impala/planner/AggregationNode.java File fe/src/main/java/org/apache/impala/planner/AggregationNode.java: http://gerrit.cloudera.org:8080/#/c/21955/13/fe/src/main/java/org/apache/impala/planner/AggregationNode.java@480 PS13, Line 480: } > This replaced taking Math.min after every multiply? I guess it caps the res Correct. saturatingMultiply() keeps multiplying and this is capping the result at aggInputCardinality. http://gerrit.cloudera.org:8080/#/c/21955/12/testdata/workloads/functional-planner/queries/PlannerTest/tpcds_cpu_cost/tpcds-q03.test File testdata/workloads/functional-planner/queries/PlannerTest/tpcds_cpu_cost/tpcds-q03.test: http://gerrit.cloudera.org:8080/#/c/21955/12/testdata/workloads/functional-planner/queries/PlannerTest/tpcds_cpu_cost/tpcds-q03.test@43 PS12, Line 43: | tuple-ids=3 row-size=52B cardinality=9.00M cost=53096454 > Is that covered by the statement After taking another look, I only find 8 planner tests that revert to pre-IMPALA-13405 state. I don't think we need to revert IMPALA-13405 entirely. Besides, other recent patch like IMPALA-13445 already have new planner tests that trigger logic from IMPALA-13445. -- To view, visit http://gerrit.cloudera.org:8080/21955 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I11f59ccc469c24c1800abaad3774c56190306944 Gerrit-Change-Number: 21955 Gerrit-PatchSet: 15 Gerrit-Owner: Riza Suminto <[email protected]> Gerrit-Reviewer: Aman Sinha <[email protected]> Gerrit-Reviewer: Impala Public Jenkins <[email protected]> Gerrit-Reviewer: Michael Smith <[email protected]> Gerrit-Reviewer: Riza Suminto <[email protected]> Gerrit-Reviewer: Yida Wu <[email protected]> Gerrit-Comment-Date: Tue, 29 Oct 2024 23:57:49 +0000 Gerrit-HasComments: Yes
