[
https://issues.apache.org/jira/browse/FLINK-36911?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17906237#comment-17906237
]
lincoln lee commented on FLINK-36911:
-------------------------------------
[~jackylau] Thank you for focusing on this part of the work, it's indeed a
potential optimization.
In terms of implementation steps, it's recommended to do the corresponding
performance evaluation, such as nexmark for streaming scenario and tpc test for
batch scenario, this can help us to decide the changes, WDYT?
> add more calcite rules to improve flink performance because calcite has
> upgrade
> -------------------------------------------------------------------------------
>
> Key: FLINK-36911
> URL: https://issues.apache.org/jira/browse/FLINK-36911
> Project: Flink
> Issue Type: Improvement
> Components: Table SQL / Planner
> Affects Versions: 2.0.0
> Reporter: Jacky Lau
> Priority: Major
> Fix For: 2.0.0
>
>
> after calcite upgrade several versions, current flink version is 2.0.0, and
> current calcite version is 1.34.
> this rules has already added more. we can align to imrpove performance
> PruneEmptyRules
>
> PruneEmptyRules.UNION_INSTANCE,
> PruneEmptyRules.INTERSECT_INSTANCE,
> PruneEmptyRules.MINUS_INSTANCE,
> PruneEmptyRules.PROJECT_INSTANCE,
> PruneEmptyRules.SORT_INSTANCE,
> PruneEmptyRules.EMPTY_TABLE_INSTANCE
> REDUCE_EXPRESSIONS
> CoreRules.WINDOW_REDUCE_EXPRESSIONS,
> CoreRules.JOIN_REDUCE_EXPRESSIONS,
> CoreRules.FILTER_VALUES_MERGE,
> CoreRules.PROJECT_FILTER_VALUES_MERGE,
> CoreRules.PROJECT_VALUES_MERGE,
> CoreRules.AGGREGATE_VALUES
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
