[
https://issues.apache.org/jira/browse/AMBARI-25523?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Zhiguo Wu resolved AMBARI-25523.
--------------------------------
Resolution: Fixed
> Ambari UI keeps loading at step3 while adding the new namespace for HDFS for
> the second time
> --------------------------------------------------------------------------------------------
>
> Key: AMBARI-25523
> URL: https://issues.apache.org/jira/browse/AMBARI-25523
> Project: Ambari
> Issue Type: Bug
> Components: ambari-web
> Affects Versions: 2.7.5
> Reporter: Akhil Naik
> Assignee: Szabolcs Beki
> Priority: Major
> Fix For: 2.8.0
>
> Time Spent: 40m
> Remaining Estimate: 0h
>
> Problem Statement : I am Trying to add namespace to HDFS via Ambari, my HDFS
> already have two namespaces : namespace1,namespace2 . when i am trying to add
> the third one say : namespace3, it fails with below script error at Step3:
> {code:java}
> app.js:5310 Uncaught TypeError: Cannot read property 'split' of undefined
> at Class.prepareDependencies (app.js:5310)
> at Class.tweakServiceConfigs (app.js:5337)
> at Class.onLoad (app.js:5293)
> at invokeAction (vendor.js:15179)
> at iterateSet (vendor.js:15161)
> at Object.sendEvent (vendor.js:15278)
> at notifyObservers (vendor.js:13870)
> at Object.Ember.notifyObservers (vendor.js:13985)
> at Object.propertyDidChange (vendor.js:14618)
> at set (vendor.js:13424)
> {code}
> Analysis : Upon analysing the code :
> https://github.com/apache/ambari/blob/release-2.7.5/ambari-web/app/controllers/main/admin/federation/step3_controller.js#L110
> It looks we are blindly taking 0th nameService as the one that have namenode1
> and namenode2 and then assiging the namenode1 and namenode2 :
> {code:java}
> var nameServices =
> App.HDFSService.find().objectAt(0).get('masterComponentGroups').mapProperty('name');
> ret.nameServicesList = nameServices.join(',');
> ret.nameservice1 = nameServices[0];
> ret.newNameservice = this.get('content.nameServiceId');
> ret.namenode1 = hdfsSiteConfigs['dfs.namenode.rpc-address.' +
> ret.nameservice1 + '.nn1'].split(':')[0];
> ret.namenode2 = hdfsSiteConfigs['dfs.namenode.rpc-address.' +
> ret.nameservice1 + '.nn2'].split(':')[0];
> {code}
> The abouve logic is wrong, when you try to fetch `nameServices` as abouve the
> order is not neccessarly the way it was created
> This logic need to rewramped.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
