squito commented on a change in pull request #23871: [SPARK-23433][SPARK-25250]
[CORE] Later created TaskSet should learn about the finished partitions
URL: https://github.com/apache/spark/pull/23871#discussion_r262561346
##########
File path: core/src/main/scala/org/apache/spark/scheduler/TaskSetManager.scala
##########
@@ -797,11 +801,12 @@ private[spark] class TaskSetManager(
maybeFinishTaskSet()
}
- private[scheduler] def markPartitionCompleted(partitionId: Int, taskInfo:
TaskInfo): Unit = {
+ private[scheduler] def markPartitionCompleted(partitionId: Int, taskInfo:
Option[TaskInfo])
+ : Unit = {
partitionToIndex.get(partitionId).foreach { index =>
if (!successful(index)) {
if (speculationEnabled && !isZombie) {
- successfulTaskDurations.insert(taskInfo.duration)
+ taskInfo.foreach { info =>
successfulTaskDurations.insert(info.duration) }
Review comment:
Yeah there was discussion about this in the past, there are arguments for
doing it multiple ways. This was kind of a compromise with something that
avoided a bug and was a reasonable change to put in, see more here:
https://github.com/apache/spark/pull/21656
----------------------------------------------------------------
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]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
