StephanEwen commented on pull request #13595:
URL: https://github.com/apache/flink/pull/13595#issuecomment-720663470
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub
StephanEwen commented on pull request #13595:
URL: https://github.com/apache/flink/pull/13595#issuecomment-718141948
Looks good form my side!
Let's wait for @zhijiangW and @gaoyunhaii and CI to give their ok as well.
We can start looking at documentation in the meantime. I
StephanEwen commented on pull request #13595:
URL: https://github.com/apache/flink/pull/13595#issuecomment-717865867
@wsry Your suggestion sounds good to me!
This is an automated message from the Apache Git Service.
To
StephanEwen commented on pull request #13595:
URL: https://github.com/apache/flink/pull/13595#issuecomment-717350591
Thanks for explaining, @wsry
Maybe we can find a good combination of "min-parallelism" and "min-buffers"
to make the switch to sort-based shuffle work well for
StephanEwen commented on pull request #13595:
URL: https://github.com/apache/flink/pull/13595#issuecomment-717204918
@wsry I have two more questions on the minimum buffers logic and confi
(1)
I previously thought that this is to ensure that the sort buffers are not
too small.
StephanEwen commented on pull request #13595:
URL: https://github.com/apache/flink/pull/13595#issuecomment-716450042
The updated logic looks good. But the name `maxBuffersPerSortMergePartition`
seems confusing. I think we should rename this to
`minBuffersPerSortMergePartition` and also
StephanEwen commented on pull request #13595:
URL: https://github.com/apache/flink/pull/13595#issuecomment-716404546
@wsry
I am confused a bit about the use of max buffers. That value is the upper
limit of buffers that will be assigned to the shuffle from the global pool. It
is