venkata91 commented on a change in pull request #34122:
URL: https://github.com/apache/spark/pull/34122#discussion_r795960916
##########
File path:
core/src/main/scala/org/apache/spark/shuffle/sort/SortShuffleManager.scala
##########
@@ -131,7 +131,7 @@ private[spark] class SortShuffleManager(conf: SparkConf)
extends ShuffleManager
metrics: ShuffleReadMetricsReporter): ShuffleReader[K, C] = {
val baseShuffleHandle = handle.asInstanceOf[BaseShuffleHandle[K, _, C]]
val (blocksByAddress, canEnableBatchFetch) =
- if (baseShuffleHandle.dependency.shuffleMergeEnabled) {
+ if (baseShuffleHandle.dependency.isShuffleMergeFinalizedMarked) {
Review comment:
@mridulm For the retry case, `shuffleMergeEnabled` would be false but
`isShuffleMergeFinalizedMarked` but we still want this to return true right to
fetch merge blocks generated by the first attempt? Therefore, won't
`isShuffleMergeFinalizedMarked` be enough? Basically as long as shuffle merge
is finalized (doesn't matter if shuffle merge is allowed or not), check for
merge blocks to be fetched.
--
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]