Ngone51 commented on a change in pull request #34156:
URL: https://github.com/apache/spark/pull/34156#discussion_r719531296
##########
File path: core/src/main/scala/org/apache/spark/MapOutputTracker.scala
##########
@@ -1448,7 +1448,7 @@ private[spark] object MapOutputTracker extends Logging {
// TODO: improve push based shuffle to read partial merged blocks
satisfying the start/end
// TODO: map indexes
if (mergeStatuses.exists(_.nonEmpty) && startMapIndex == 0
- && endMapIndex == mapStatuses.length) {
+ && endMapIndex == mapStatuses.length && endPartition - startPartition ==
1) {
Review comment:
> My concern with only keeping this condition (endPartition -
startPartition == 1) instead of disable batch fetch is that it will effectively
disable push-based shuffle in 3.2.0 even if the push.enabled flag is set to
true.
~~IUC, push-based shuffle currently can only work with one reduce partition,
no? ~~ And I didn't see `endPartition - startPartition == 1` disabled
push-based shuffle (PBS) for any cases except "partition coalesce".
To clarify, there's no other cases would have multiple reduce partition if
PBS wants to take effect due to the limit of `startMapIndex == 0 &&
endMapIndex == mapStatuses.length`.
--
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]
