[
https://issues.apache.org/jira/browse/SPARK-2666?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14725615#comment-14725615
]
Imran Rashid commented on SPARK-2666:
-------------------------------------
I realized I didn't very clearly spell out one of my main points: I am
proposing widening this issue to not be about only {{FetchFailed}}. I think
instead we should consider changing this issue to refactor the code to unify
"zombification" and cancelling tasks. In general I know that smaller changes
are better, especially related to the scheduler, but in this case I think we'll
be able to improve the code by tackling them together.
> when task is FetchFailed cancel running tasks of failedStage
> ------------------------------------------------------------
>
> Key: SPARK-2666
> URL: https://issues.apache.org/jira/browse/SPARK-2666
> Project: Spark
> Issue Type: Bug
> Components: Spark Core
> Reporter: Lianhui Wang
>
> in DAGScheduler's handleTaskCompletion,when reason of failed task is
> FetchFailed, cancel running tasks of failedStage before add failedStage to
> failedStages queue.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
