Ngone51 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_r246616157
##########
File path: core/src/main/scala/org/apache/spark/scheduler/DAGScheduler.scala
##########
@@ -1383,6 +1383,8 @@ private[spark] class DAGScheduler(
if (!job.finished(rt.outputId)) {
job.finished(rt.outputId) = true
job.numFinished += 1
+
taskScheduler.markPartitionIdAsCompletedAndKillCorrespondingTaskAttempts(
Review comment:
Yeah, I agree this could be a extremely rare case and your work has proved
that this pr is really a effective fix.
And as I mentioned above, if we could record this completed partition in
`TaskScheduler` at this point and telling others TaskSets about the completed
patitions once they're created in `TaskScheduler`, we may avoid the potential
issue totally. But, I think we don't need to do this right now since your fix
has already proved to be effective. Instead, how about leaving some comments to
explain the potential issue ? Thus, we may easily fix it once we really hit it.
----------------------------------------------------------------
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]
