You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Marcelo Vanzin (JIRA)" <ji...@apache.org> on 2017/10/13 17:16:00 UTC

[jira] [Commented] (SPARK-22275) ReplayListenerSuite doesn't clean up properly after itself

    [ https://issues.apache.org/jira/browse/SPARK-22275?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16203887#comment-16203887 ] 

Marcelo Vanzin commented on SPARK-22275:
----------------------------------------

Actually this is caused by the exception handling in SparkContext:

{noformat}
  } catch {
    case NonFatal(e) =>
{noformat}

The test is throwing a {{java.lang.Error}} so it's not caught and the context is left alive... it sounds prudent to at least try to clean up when those occur too.

> ReplayListenerSuite doesn't clean up properly after itself
> ----------------------------------------------------------
>
>                 Key: SPARK-22275
>                 URL: https://issues.apache.org/jira/browse/SPARK-22275
>             Project: Spark
>          Issue Type: Bug
>          Components: Tests
>    Affects Versions: 2.3.0
>            Reporter: Marcelo Vanzin
>            Priority: Minor
>
> The test failures from https://github.com/apache/spark/pull/18805 show a bunch of errors in the logs related to this suite:
> {noformat}
> 17/10/13 06:53:26.609 dispatcher-event-loop-15 ERROR Worker: Failed to launch executor app-20171013030138-0000/3 for Test replay.
> java.lang.IllegalStateException: Shutdown hooks cannot be modified during shutdown.
>         at org.apache.spark.util.SparkShutdownHookManager.add(ShutdownHookManager.scala:195)
> {noformat}
> That's the app name for this test but the error message shows up in logs in a completely separate test suite.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org