andygrove commented on pull request #32195:
URL: https://github.com/apache/spark/pull/32195#issuecomment-820888554


   > Does it mean, if Spark converts reused query stage from columnar to 
row-based, it will be compatible? Just out of curiosity.
   
   @viirya Spark plugins can create either columnar or row-based exchanges as 
required for compatibility with other parts of the query plan. 
   
   For example. a columnar `BroadcastHashJoin` would expect both of its child 
plans to also be columnar, so if Spark replaces one of the child plans with a 
row-based plan it would not be compatible at execution time.


-- 
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]

Reply via email to