Ngone51 commented on a change in pull request #30164:
URL: https://github.com/apache/spark/pull/30164#discussion_r522563570
##########
File path: core/src/main/scala/org/apache/spark/scheduler/TaskSetManager.scala
##########
@@ -824,6 +824,8 @@ private[spark] class TaskSetManager(
if (fetchFailed.bmAddress != null) {
healthTracker.foreach(_.updateExcludedForFetchFailure(
fetchFailed.bmAddress.host, fetchFailed.bmAddress.executorId))
+ // Remove fetchFailed host in the shuffle push merger list for push
based shuffle
+
env.blockManager.master.removeShufflePushMergerLocation(fetchFailed.bmAddress.host)
Review comment:
hmm..I'd prefer to handle this inside DAGScheduler...DAGScheduler
handles fetch failure too and it also checks whether external shuffle service
is used in case of the failure, see:
https://github.com/apache/spark/blob/72ad9dcd5d484a8dd64c08889de85ef9de2a6077/core/src/main/scala/org/apache/spark/scheduler/DAGScheduler.scala#L1826
----------------------------------------------------------------
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:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
