[
https://issues.apache.org/jira/browse/SPARK-8562?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Apache Spark reassigned SPARK-8562:
-----------------------------------
Assignee: Apache Spark
> Annoying messages about executor lost after stopping SparkContext
> -----------------------------------------------------------------
>
> Key: SPARK-8562
> URL: https://issues.apache.org/jira/browse/SPARK-8562
> Project: Spark
> Issue Type: Bug
> Components: Spark Core
> Affects Versions: 1.3.1
> Reporter: Jacek Lewandowski
> Assignee: Apache Spark
> Priority: Minor
>
> This is a kind of a race condition because when the context is being stopped,
> the executors are stopped first and then the actor system. However, sometimes
> it happens that the actor system receives the information about stopping
> executors before it gets shutdown and therefore it reports the "Executor
> lost" error.
> I think the simple way to fix that would be to log errors in
> {{TaskSchedulerImpl.executorLost}} method, only if Spark context isn't
> stopped (stopped flag is set to true first).
--
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
