hadoop-yarn-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Robert Kanter (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (YARN-5876) TestResourceTrackerService#testGracefulDecommissionWithApp fails intermittently on trunk
Date Wed, 21 Jun 2017 23:19:00 GMT

     [ https://issues.apache.org/jira/browse/YARN-5876?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel

Robert Kanter updated YARN-5876:
    Attachment: YARN-5876.001.patch

This was a hard one to reproduce - I was only able to by inserting some sleeps in the right
places.  It turns out that there's a very small period of time while a decommissioning node
is being moved from {{nodes}} to {{inactiveNodes}} where the node is in neither map (because
it's being moved).  If the timing works out just perfectly, {{MockRM}} tries to get the node
during that window, and can't find it, resulting in {{null}}.

The patch fixes the problem by including retries for finding the node as part of the timeout
in {{MockRM#waitForState}}, instead of the original code that only tried once.

> TestResourceTrackerService#testGracefulDecommissionWithApp fails intermittently on trunk
> ----------------------------------------------------------------------------------------
>                 Key: YARN-5876
>                 URL: https://issues.apache.org/jira/browse/YARN-5876
>             Project: Hadoop YARN
>          Issue Type: Bug
>            Reporter: Varun Saxena
>            Assignee: Robert Kanter
>         Attachments: YARN-5876.001.patch
> {noformat}
> java.lang.AssertionError: node shouldn't be null
> 	at org.junit.Assert.fail(Assert.java:88)
> 	at org.junit.Assert.assertTrue(Assert.java:41)
> 	at org.junit.Assert.assertNotNull(Assert.java:621)
> 	at org.apache.hadoop.yarn.server.resourcemanager.MockRM.waitForState(MockRM.java:750)
> 	at org.apache.hadoop.yarn.server.resourcemanager.TestResourceTrackerService.testGracefulDecommissionWithApp(TestResourceTrackerService.java:318)
> {noformat}
> Refer to https://builds.apache.org/job/PreCommit-YARN-Build/13884/testReport/

This message was sent by Atlassian JIRA

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

View raw message