[jira] [Commented] (YARN-3839) Quit throwing NMNotYetReadyException

["Manikandan R (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/YARN-3839?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15996745#comment-15996745 ] 

Manikandan R commented on YARN-3839:
------------------------------------

{quote}I don't believe we should remove the NMNotYetReadyException from the ServeryProxy exception map. Otherwise if a newer client talks to an older server that is still throwing the old exception then the new client will not handle it gracefully.{quote}
Ok. Taken care. Attached new patch.
{quote}Please look a bit deeper into the test failures, there may be races that are not going well once the check for the blocked boolean was removed and now we're proceeding quicker than we would have before.{quote}
I spent some time on these failures, but don't see any issues. For example, {code}java.lang.Exception: test timed out after 120000 milliseconds{code} failures in TestContainerManagerSecurity.testContainerManager:158->testNMTokens:253->testStartContainer:478->startContainer:557 is reproducible even in my local working copy and can be avoided by increasing timeout value for specific test case. However, am getting NPE in line containing {code}nmContet.getContainers().get(containerId){code} in {code}waitForContainerToFinishOnNM(){code} method in TestContainerManagerSecurity.java, but don't think it is related to this patch. Please let me know your thoughts.

> Quit throwing NMNotYetReadyException
> ------------------------------------
>
>                 Key: YARN-3839
>                 URL: https://issues.apache.org/jira/browse/YARN-3839
>             Project: Hadoop YARN
>          Issue Type: Bug
>          Components: nodemanager
>            Reporter: Karthik Kambatla
>            Assignee: Manikandan R
>         Attachments: YARN-3839.001.patch, YARN-3839.002.patch, YARN-3839.003.patch, YARN-3839.004.patch, YARN-3839.005.patch, YARN-3839.006.patch, YARN-3839.007.patch
>
>
> Quit throwing NMNotYetReadyException when NM has not yet registered with the RM.



--
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