[
https://issues.apache.org/jira/browse/SPARK-34931?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dongjoon Hyun updated SPARK-34931:
----------------------------------
Description: when executor lost for some reason(e.g. Unable to register
with external shuffle server)，CoarseGrainedExecutorBackend will send a
RemoveExecutor event with 'ExecutorLossReason'. But this will cause
TaskSetManager handle handleFailedTask function with exitCausedByApp=true.
This is not correct (was: {code}
The downloaded source packages are in
'/tmp/RtmpgmgGfi/downloaded_packages'
Using bundled GitHub PAT. Please add your own PAT to the env var `GITHUB_PAT`
Downloading GitHub repo jimhester/[email protected]
/bin/tar: This does not look like a tar archive
{code}
https://github.com/apache/spark/pull/32018/checks?check_run_id=2242498794)
> CoarseGrainedExecutorBackend send wrong 'Reason' when executor exits which
> leading to job failed.
> -------------------------------------------------------------------------------------------------
>
> Key: SPARK-34931
> URL: https://issues.apache.org/jira/browse/SPARK-34931
> Project: Spark
> Issue Type: Test
> Components: Project Infra
> Affects Versions: 2.4.7, 3.0.2, 3.2.0, 3.1.1
> Reporter: Hyukjin Kwon
> Priority: Major
>
> when executor lost for some reason(e.g. Unable to register with external
> shuffle server)，CoarseGrainedExecutorBackend will send a RemoveExecutor event
> with 'ExecutorLossReason'. But this will cause TaskSetManager handle
> handleFailedTask function with exitCausedByApp=true. This is not correct
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
