[
https://issues.apache.org/jira/browse/FLINK-19822?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17609812#comment-17609812
]
Alexander Smirnov commented on FLINK-19822:
-------------------------------------------
Hi [~godfreyhe]! I'm wondering why this improvement is still open, although it
looks quite useful. Are there any pitfalls in implementation from linked
Github PR?
> Remove redundant shuffle for streaming
> --------------------------------------
>
> Key: FLINK-19822
> URL: https://issues.apache.org/jira/browse/FLINK-19822
> Project: Flink
> Issue Type: Improvement
> Components: Table SQL / Planner
> Reporter: godfrey he
> Priority: Not a Priority
> Labels: auto-deprioritized-major, auto-deprioritized-minor,
> auto-unassigned, pull-request-available
>
> This is similar
> [FLINK-12575|https://issues.apache.org/jira/browse/FLINK-12575], we could
> implement {{satisfyTraits}} method for stream nodes to remove redundant
> shuffle. This could add more possibilities that more operators can be merged
> into multiple input operator.
> Different batch, stream operators require the shuffle keys and the state keys
> must be exactly the same, otherwise the state may be not correct.
> We only support a few operators in this issue, such as Join and regular
> Aggregate. Other operators will be supported in the future.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
