kgyrtkirk commented on a change in pull request #1439:
URL: https://github.com/apache/hive/pull/1439#discussion_r482092255
##########
File path:
ql/src/java/org/apache/hadoop/hive/ql/optimizer/calcite/rules/HiveAggregateJoinTransposeRule.java
##########
@@ -290,7 +291,8 @@ public void onMatch(RelOptRuleCall call) {
RelNode r = relBuilder.build();
RelOptCost afterCost = mq.getCumulativeCost(r);
RelOptCost beforeCost = mq.getCumulativeCost(aggregate);
- if (afterCost.isLt(beforeCost)) {
Review comment:
yes; if we use `isLe` the current cost model which only takes rowcount
into account will prefer the pushing aggregates further
there is another alternative to the `force` based approach: the rule can be
configured to use the more advanced cost system - so that it could take cpu/io
cost into account
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
