hadoop-yarn-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Yufei Gu (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (YARN-6326) Shouldn't use AppAttemptIds to fetch applications while AM Simulator tracks app in SLS
Date Sat, 18 Mar 2017 06:55:41 GMT

    [ https://issues.apache.org/jira/browse/YARN-6326?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15931095#comment-15931095
] 

Yufei Gu commented on YARN-6326:
--------------------------------

Thanks [~rkanter] for the review. Uploaded patch v5 for your comments.
# Fixed
# I don't think it is a good idea to add a method into both interface after offline discussions.
I am not fan of current design of {{YarnScheduler}} and {{ResourceScheduerl}}. But I realize
that we'd better fix it in another Jira if there is any issue in them. Comparing to incompatibility,
downcast is not so terrible. So I downcast {{ResourceScheduler}} to {{AbstractYarnScheduler}}
to get the method I need.

> Shouldn't use AppAttemptIds to fetch applications while AM Simulator tracks app in SLS
> --------------------------------------------------------------------------------------
>
>                 Key: YARN-6326
>                 URL: https://issues.apache.org/jira/browse/YARN-6326
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: scheduler-load-simulator
>            Reporter: Yufei Gu
>            Assignee: Yufei Gu
>         Attachments: YARN-6326.001.patch, YARN-6326.002.patch, YARN-6326.003.patch, YARN-6326.004.patch,
YARN-6326.005.patch
>
>
> This causes a NPE issue. Beside the NPE, the metrics won't reflect the different attempts.
We should pass ApplicationId Instead of AppAttemptId. The NPE caused by the issue:
> {code}
> 2017-03-13 20:43:39,153 INFO appmaster.AMSimulator: Submit a new application application_1489463017173_0001
> java.lang.NullPointerException
> 	at org.apache.hadoop.yarn.server.resourcemanager.scheduler.AbstractYarnScheduler.getApplicationAttempt(AbstractYarnScheduler.java:327)
> 	at org.apache.hadoop.yarn.server.resourcemanager.scheduler.fair.FairScheduler.getSchedulerApp(FairScheduler.java:1028)
> 	at org.apache.hadoop.yarn.sls.scheduler.FairSchedulerMetrics.trackApp(FairSchedulerMetrics.java:68)
> 	at org.apache.hadoop.yarn.sls.scheduler.ResourceSchedulerWrapper.addTrackedApp(ResourceSchedulerWrapper.java:799)
> 	at org.apache.hadoop.yarn.sls.appmaster.AMSimulator.trackApp(AMSimulator.java:338)
> 	at org.apache.hadoop.yarn.sls.appmaster.AMSimulator.firstStep(AMSimulator.java:156)
> 	at org.apache.hadoop.yarn.sls.scheduler.TaskRunner$Task.run(TaskRunner.java:90)
> 	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
> 	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
> 	at java.lang.Thread.run(Thread.java:745)
> Exception in thread "pool-6-thread-1" java.lang.NullPointerException
> 	at org.apache.hadoop.yarn.sls.scheduler.TaskRunner$Task.run(TaskRunner.java:105)
> 	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
> 	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
> 	at java.lang.Thread.run(Thread.java:745)
> {code}



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

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


Mime
View raw message