[
https://issues.apache.org/jira/browse/AMBARI-14664?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15099061#comment-15099061
]
Yusaku Sako commented on AMBARI-14664:
--------------------------------------
UT failures seem unrelated to this patch. Most likely AMBARI-14670
> Blueprints: After Kerberos deployment, UI requires that Yarn service be
> restarted
> ---------------------------------------------------------------------------------
>
> Key: AMBARI-14664
> URL: https://issues.apache.org/jira/browse/AMBARI-14664
> Project: Ambari
> Issue Type: Bug
> Components: ambari-server
> Affects Versions: 2.2.0
> Reporter: Sandor Magyari
> Assignee: Sandor Magyari
> Priority: Critical
> Fix For: 2.2.1
>
> Attachments: AMBARI-14664.patch
>
>
> This is similar issue to AMBARI-14251 Stale configuration problem.
> Kerberos.json config files contains a few properties like:
> {$clusterHostInfo/webhcat_server_host} for HIVE and KNOX configs. These
> properties are replaced using StageUtils.getClusterHostInfo() which is a
> quite problematic one and sometimes could happen that hosts are missing from
> clusterHostInfo since TopologyHostRequest were not created at that time.
> The solution would be to pass clusterHostInfo map to KerberosHelper in case
> of using it from TopologyManager so there's no need to get clusterInfo from
> StageUtils. We can create this map from ClusterTopology once topology is
> resolved. To retrieve hosts to be resolved form kerberos configs, we would
> pass an empty componentHostMap then we can select host variables from config
> properties. This way the same Kerberos config logic will run both for BP and
> non BP deployments.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
