[
https://issues.apache.org/jira/browse/AMBARI-12910?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14724324#comment-14724324
]
Hudson commented on AMBARI-12910:
---------------------------------
FAILURE: Integrated in Ambari-trunk-Commit #3357 (See
[https://builds.apache.org/job/Ambari-trunk-Commit/3357/])
AMBARI-12910. Ambari Views Framework - need support for accessing RM running
under HA. (Dipayan Bhowmick via yusaku) (yusaku:
http://git-wip-us.apache.org/repos/asf?p=ambari.git&a=commit&h=ba04e8df9385a3171adf5c3e296cf3782c6b93cc)
* contrib/views/hive/src/main/resources/view.xml
*
contrib/views/utils/src/test/java/org/apache/ambari/view/utils/ambari/ServicesTest.java
*
contrib/views/utils/src/main/java/org/apache/ambari/view/utils/ambari/Services.java
> Ambari Views Framework - need support for accessing RM running under HA
> -----------------------------------------------------------------------
>
> Key: AMBARI-12910
> URL: https://issues.apache.org/jira/browse/AMBARI-12910
> Project: Ambari
> Issue Type: Bug
> Components: ambari-views
> Affects Versions: 2.1.0
> Reporter: DIPAYAN BHOWMICK
> Assignee: DIPAYAN BHOWMICK
> Fix For: 2.1.2
>
> Attachments: AMBARI-12910_branch-2.1.patch
>
> Original Estimate: 168h
> Remaining Estimate: 168h
>
> Currently all views like the Tez view, YARN cluster view and YARN admin views
> that need to access YARN RMs running HA don't handle this scenario correctly.
> We need support in the views framework to enable this (note, if both RMs are
> running the RMs handle redirecting from the standby RM to the active one, but
> it has to be tested to work with the Ambari proxy. If one of the RMs is
> actually down, things are broken today).
> The Hive view might also need this.
> The following changes need to be considered:
> 1. The property that used to specify the RM location has to be extended to
> take a list of RM locations
> 2. The Ambari proxy has to cycle through the locations until it find one that
> works
> 3. For auto-created view instances, based on the cluster configuration, the
> multi-RM property in (1) has to be set correctly without user intervention
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
