[
https://issues.apache.org/jira/browse/AMBARI-25164?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
ASF GitHub Bot updated AMBARI-25164:
------------------------------------
Labels: pull-request-available (was: )
> Rack "Config Refresh" behaviour is different in Ambari 2.6 and 2.7.3
> --------------------------------------------------------------------
>
> Key: AMBARI-25164
> URL: https://issues.apache.org/jira/browse/AMBARI-25164
> Project: Ambari
> Issue Type: Bug
> Components: ambari-web
> Affects Versions: 2.7.3
> Reporter: Andrii Tkach
> Assignee: Andrii Tkach
> Priority: Major
> Labels: pull-request-available
> Fix For: 2.7.4
>
>
> In the new ambari-2.7.3 version, there is no way (via web ui) to generally
> refresh configs on a host . There is a way to refresh HDFS_CLIENT config on
> the Namenode host, but refreshing HDFS_CLIENT config *does not* refresh the
> topology_mappings.data file.
> *In Ambari 2.6.2: (During the "Refresh Configs")*
> When we refresh "Refresh Configs" then it refreshes all client configs which
> also includes HDFS and MapReduce2 config refresh. During the "Restart
> MapReduce2 Client" step it actually performs the
> "/etc/hadoop/conf/topology_mappings.data" file refresh with correct rack info
> which is updated via ambari ui.
> Example :
> {code}
> 2019-02-09 21:31:05,642 -
> File['/usr/hdp/2.6.5.0-292/hadoop/conf/configuration.xsl'] {'owner': 'hdfs',
> 'group': 'hadoop'}
> 2019-02-09 21:31:05,647 - File['/etc/hadoop/conf/topology_mappings.data']
> {'owner': 'hdfs', 'content': Template('topology_mappings.data.j2'),
> 'only_if': 'test -d /etc/hadoop/conf', 'group': 'hadoop', 'mode': 0644}
> 2019-02-09 21:31:05,655 - File['/etc/hadoop/conf/topology_script.py']
> {'content': StaticFile('[email protected]'), 'only_if': 'test -d /etc/hadoop/conf',
> 'mode': 0755}
> {code}
> *In Ambari 2.7.3: (During the "Refresh Configs")*
> Even after refreshing HDFS Client (AND) MapReduce Clients individually using
> "Refresh Configs" iwe do not see above kind of message. Also no changes
> happens inside the "/etc/hadoop/conf/topology_mappings.data" is not updated
> until we restart daemon components like DataNode. So the behavior of client
> config refresh seems to be slightly changed from ambari side.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
