Github user scwf commented on the issue:
https://github.com/apache/spark/pull/16633
Your proposal avoid the cost of all partitions compute and shuffle for
local limit but introduce some partitions recompute for local limit stage.
We can not decide which cost is cheaper(in most cases), note computation
logical for local limit stage maybe very complex and costly.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
