hadoop-yarn-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Xuan Gong (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (YARN-1811) Error 500 when clicking the "Application Master" link in the RM UI while a job is running with RM HA
Date Mon, 10 Mar 2014 22:20:45 GMT

    [ https://issues.apache.org/jira/browse/YARN-1811?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13926333#comment-13926333
] 

Xuan Gong commented on YARN-1811:
---------------------------------

[~kkambatl] [~rkanter]  I still do not understand why this happens.
No matter we start proxy server separately or embedded with RM, we have already added retry
logic there. If we click the "Application Master" link in RM UI, even if the RM becomes standby,
it will do the re-try until it finds the active RM, and will redirect to the right url.

Also, i am not quite sure why the patch can fix the issue. AmIpFilter calls WebAppUtils.getProxyHostAndPort
only when it does initialization. At that time, it can get the active RM's address. But if
failover happens after that, the AmIpFilter still does not know the current active RM. So,
how does this fix the issue ?

> Error 500 when clicking the "Application Master" link in the RM UI while a job is running
with RM HA
> ----------------------------------------------------------------------------------------------------
>
>                 Key: YARN-1811
>                 URL: https://issues.apache.org/jira/browse/YARN-1811
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: resourcemanager
>    Affects Versions: 2.3.0
>            Reporter: Robert Kanter
>            Assignee: Robert Kanter
>         Attachments: YARN-1811.patch
>
>
> When using RM HA, if you click on the "Application Master" link in the RM web UI while
the job is running, you get an Error 500:



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Mime
View raw message