[jira] [Commented] (AMBARI-5032) YARN Scheduler section drop lines when saving a scheduler configuration

["Rémy SAISSY (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/AMBARI-5032?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13936323#comment-13936323 ] 

Rémy SAISSY commented on AMBARI-5032:
-------------------------------------

I digged a bit into the issue and found out what happens

Given this default configuration (of Ambari-1.4.1 with the HDP2.0.6 stack):
yarn.scheduler.capacity.maximum-am-resource-percent=0.2
yarn.scheduler.capacity.maximum-applications=10000
yarn.scheduler.capacity.root.acl_administer_queues=*
yarn.scheduler.capacity.root.capacity=100
yarn.scheduler.capacity.root.default.acl_administer_jobs=*
yarn.scheduler.capacity.root.default.acl_submit_jobs=*
yarn.scheduler.capacity.root.default.capacity=100
yarn.scheduler.capacity.root.default.maximum-capacity=100
yarn.scheduler.capacity.root.default.state=RUNNING
yarn.scheduler.capacity.root.default.user-limit-factor=1
yarn.scheduler.capacity.root.queues=default
yarn.scheduler.capacity.root.unfunded.capacity=50

Changing:
yarn.scheduler.capacity.root.default.acl_administer_jobs=*

to:
yarn.scheduler.capacity.root.default.acl_administer_jobs=user1,user2

works as expected.

However, changing it to:
yarn.scheduler.capacity.root.default.acl_administer_jobs=user1 group1

saves only user1 as the value. Everything starting from the white space is ignored.

It happens that in config.js in the web client:
  createSiteObj: function (siteName, tagName) {
    var siteObj = this.get('uiConfigs').filterProperty('filename', siteName + '.xml');

At this point, the siteObj property have already been cleaned from the part of the value after the whitespace.
I don't know where the cleaning and the setting of uiConfigs happens.


> YARN Scheduler section drop lines when saving a scheduler configuration
> -----------------------------------------------------------------------
>
>                 Key: AMBARI-5032
>                 URL: https://issues.apache.org/jira/browse/AMBARI-5032
>             Project: Ambari
>          Issue Type: Bug
>    Affects Versions: 1.4.1
>            Reporter: Rémy SAISSY
>
> Hi,
> in Ambari, Services:YARN::config::Scheduler, when I copy paste the following configuration:
> yarn.scheduler.capacity.maximum-am-resource-percent=0.2
> yarn.scheduler.capacity.maximum-applications=10000
> yarn.scheduler.capacity.root.capacity=100
> yarn.scheduler.capacity.root.queues=default,datalab,batch,restit
> yarn.scheduler.capacity.root.default.capacity=5
> yarn.scheduler.capacity.root.default.maximum-capacity=5
> yarn.scheduler.capacity.root.default.state=RUNNING
> yarn.scheduler.capacity.root.default.acl_administer_queue= users
> yarn.scheduler.capacity.root.datalab.capacity=10
> yarn.scheduler.capacity.root.datalab.maximum-capacity=10
> yarn.scheduler.capacity.root.datalab.state=RUNNING
> yarn.scheduler.capacity.root.datalab.acl_submit_applications= datalab_users
> yarn.scheduler.capacity.root.datalab.acl_administer_queue= datalab_admins
> yarn.scheduler.capacity.root.batch.capacity=40
> yarn.scheduler.capacity.root.batch.maximum-capacity=40
> yarn.scheduler.capacity.root.batch.state=RUNNING
> yarn.scheduler.capacity.root.batch.acl_submit_applications= batch_users
> yarn.scheduler.capacity.root.batch.acl_administer_queue= batch_admins
> yarn.scheduler.capacity.root.default.capacity=45
> yarn.scheduler.capacity.root.default.maximum-capacity=45
> yarn.scheduler.capacity.root.default.state=RUNNING
> yarn.scheduler.capacity.root.default.acl_submit_applications= restit_users
> yarn.scheduler.capacity.root.default.acl_administer_queue= restit_admins
> and click Save, Ambari saves the following invalid configuration:
> yarn.scheduler.capacity.maximum-am-resource-percent=0.2
> yarn.scheduler.capacity.maximum-applications=10000
> yarn.scheduler.capacity.root.batch.capacity=40
> yarn.scheduler.capacity.root.batch.maximum-capacity=40
> yarn.scheduler.capacity.root.batch.state=RUNNING
> yarn.scheduler.capacity.root.capacity=100
> yarn.scheduler.capacity.root.datalab.capacity=10
> yarn.scheduler.capacity.root.datalab.maximum-capacity=10
> yarn.scheduler.capacity.root.datalab.state=RUNNING
> yarn.scheduler.capacity.root.default.capacity=45
> yarn.scheduler.capacity.root.default.maximum-capacity=45
> yarn.scheduler.capacity.root.default.state=RUNNING
> yarn.scheduler.capacity.root.queues=default,datalab,batch,restit
> And the ResourceManager doesn't restart properly due to a Capacity Scheduler configuration file issue.



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