c21 edited a comment on pull request #35552:
URL: https://github.com/apache/spark/pull/35552#issuecomment-1045066962


   > I suspect partial agg is not main problem in your internal prod setting as 
per https://github.com/apache/spark/pull/28804#issuecomment-854089520, but the 
parallelism of final agg is, correct?
   
   @sigmod - I would like to elaborate more about where I come from. I am not 
pushing back the feature with the goal to work with my internal company 
environment. In Spark community, we are trying to move forward with [SPIP - 
storage partitioned join](https://issues.apache.org/jira/browse/SPARK-37375), 
and deprecate `HashClusteredDistribution` is one part of work. So it's not 
preferred to add the whole thing back if there's alternative option. For 
stream-stream join operator, we have to add the distribution 
`StatefulOpClusteredDistribution` back for state store correctness only, and I 
don't think it's in the right direction to make it more general. I thought we 
only wanted to make a special case for SS operators, but look like here we want 
to keep expanding the functionality.


-- 
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.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]



---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to