mridulm edited a comment on pull request #30650: URL: https://github.com/apache/spark/pull/30650#issuecomment-771271706
I am fine with preserving the behavior - but it looks broken to me. Just to clarify the behavior as I understand - * If there is insufficient resources to run the entire taskset at (some) highest locality level, and there is atleast one task that can be launched at that level - we will end up always failing the barrier taksset (unless delay is set to 0) ? If yes, as a follow up, do we want to make delay as 0 for barrier tasksets ? And if delay is 0, I am assuming the benefits of this PR will apply to barrier tasksets as well (in legacy) ? Thoughts ? ---------------------------------------------------------------- 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: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org