squito commented on a change in pull request #22806: [SPARK-25250][CORE] : Late
zombie task completions handled correctly even before new taskset launched
URL: https://github.com/apache/spark/pull/22806#discussion_r250363619
##########
File path:
core/src/main/scala/org/apache/spark/scheduler/TaskSchedulerImpl.scala
##########
@@ -287,6 +287,17 @@ private[spark] class TaskSchedulerImpl(
}
}
+ override def completeTasks(partitionId: Int, stageId: Int, taskInfo:
TaskInfo): Unit = {
+ taskSetsByStageIdAndAttempt.getOrElse(stageId, Map()).values.foreach { tsm
=>
+ tsm.partitionToIndex.get(partitionId) match {
+ case Some(index) =>
+ tsm.markPartitionCompleted(index, taskInfo)
Review comment:
partition and index seem mixed up here. They are usually the same on the
first stage attempt, but are almost always different later on, so they need to
be treated with care. `tsm.markPartitionCompleted` expects a partition
currently. I think you should just skip doing the
`tsm.partitionToIndex.get(partitionId)` here.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
