[
https://issues.apache.org/jira/browse/SPARK-3469?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14128149#comment-14128149
]
Apache Spark commented on SPARK-3469:
-------------------------------------
User 'rxin' has created a pull request for this issue:
https://github.com/apache/spark/pull/2343
> All TaskCompletionListeners should be called even if some of them fail
> ----------------------------------------------------------------------
>
> Key: SPARK-3469
> URL: https://issues.apache.org/jira/browse/SPARK-3469
> Project: Spark
> Issue Type: Improvement
> Components: Spark Core
> Affects Versions: 1.1.0
> Reporter: Reynold Xin
> Assignee: Reynold Xin
>
> If there are multiple TaskCompletionListeners, and any one of them misbehaves
> (e.g. throws an exception), then we will skip executing the rest of them.
> As we are increasingly relying on TaskCompletionListener for cleaning up of
> resources, we should make sure they are always called, even if the previous
> ones fail.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
