[
https://issues.apache.org/jira/browse/YARN-4380?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15024872#comment-15024872
]
Varun Saxena commented on YARN-4380:
------------------------------------
[~ozawa], checked the test failure.
This issue is coming because we do not have any wait after sending
DESTROY_APPLICATION_RESOURCES event. And we are immediately checking for
APPLICATION_RESOURCES_CLEANEDUP event being sent by Resource Localization
service(upon processing of DESTROY_APPLICATION_RESOURCES).
{code}
LocalizationEvent destroyApp =
new ApplicationLocalizationEvent(
LocalizationEventType.DESTROY_APPLICATION_RESOURCES, app);
spyService.handle(destroyApp);
verify(applicationBus).handle(argThat(matchesAppDestroy));
{code}
Adding a {{dispatcher.await()}} statement in between spyService.handle and
verify statement will resolve the issue.
This test case is broken by YARN-90 and not YARN-2902. So we should raise
another JIRA for it. As you found the issue, you can raise the JIRA and assign
it to me. I will update a patch there.
> TestResourceLocalizationService.testDownloadingResourcesOnContainerKill fails
> intermittently on branch-2.8
> ----------------------------------------------------------------------------------------------------------
>
> Key: YARN-4380
> URL: https://issues.apache.org/jira/browse/YARN-4380
> Project: Hadoop YARN
> Issue Type: Bug
> Components: test
> Affects Versions: 2.8.0, 2.7.1
> Reporter: Tsuyoshi Ozawa
> Assignee: Varun Saxena
> Attachments: YARN-4380.01.patch,
> org.apache.hadoop.yarn.applications.distributedshell.TestDistributedShell-output.2.txt,
>
> org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.TestResourceLocalizationService-output.txt
>
>
> {quote}
> Tests run: 14, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 5.361 sec
> <<< FAILURE! - in
> org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.TestResourceLocalizationService
> testDownloadingResourcesOnContainerKill(org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.TestResourceLocalizationService)
> Time elapsed: 0.109 sec <<< FAILURE!
> org.mockito.exceptions.verification.junit.ArgumentsAreDifferent:
> Argument(s) are different! Wanted:
> deletionService.delete(
> "user0",
> null,
> <Downloading paths matcher>
> );
> -> at
> org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.TestResourceLocalizationService.testDownloadingResourcesOnContainerKill(TestResourceLocalizationService.java:1322)
> Actual invocation has different arguments:
> deletionService.delete(
> "user0",
>
> /home/ubuntu/hadoop-dev/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/target/org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.TestResourceLocalizationService/0/usercache/user0/appcache/application_314159265358979_0003/container_314159265358979_0003_01_000042
> );
> -> at
> org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.TestResourceLocalizationService.testDownloadingResourcesOnContainerKill(TestResourceLocalizationService.java:1296)
> at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native
> Method)
> at
> sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:62)
> at
> sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)
> at java.lang.reflect.Constructor.newInstance(Constructor.java:422)
> at
> org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.TestResourceLocalizationService.testDownloadingResourcesOnContainerKill(TestResourceLocalizationService.java:1322)
> {quote}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
