You are viewing a plain text version of this content. The canonical link for it is here.
Posted to yarn-issues@hadoop.apache.org by "Szilard Nemeth (Jira)" <ji...@apache.org> on 2019/10/31 12:17:00 UTC
[jira] [Commented] (YARN-9780) SchedulerConf Mutation API does not
Allow Stop and Remove Queue in a single call
[ https://issues.apache.org/jira/browse/YARN-9780?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16963935#comment-16963935 ]
Szilard Nemeth commented on YARN-9780:
--------------------------------------
Hi [~prabhujoseph]!
Thanks for this patch!
Looks good in overall, but one thing:
1. Don't we need to update the javadoc of method org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacitySchedulerQueueManager#validateQueueHierarchy?
Currently, it says:
{code:java}
Ensure all existing queues are present. Queues cannot be deleted if its not
* in Stopped state, Queue's cannot be moved from one hierarchy to other also.
* Previous child queue could be converted into parent queue if it is in
* STOPPED state.
{code}
This statement is not true anymore: Queues cannot be deleted if its not in Stopped state.
If you are already editing the javadoc, please make sure the state names are consistent (STOPPED vs. Stopped) and fix the "its" to "it is".
Thanks!
> SchedulerConf Mutation API does not Allow Stop and Remove Queue in a single call
> --------------------------------------------------------------------------------
>
> Key: YARN-9780
> URL: https://issues.apache.org/jira/browse/YARN-9780
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: capacity scheduler
> Affects Versions: 3.3.0
> Reporter: Prabhu Joseph
> Assignee: Prabhu Joseph
> Priority: Major
> Attachments: YARN-9780-001.patch, YARN-9780-002.patch, YARN-9780-003.patch, YARN-9780-004.patch
>
>
> SchedulerConf Mutation API does not Allow Stop and Remove Queue in a single call. The queue has to be stopped before removing and so it is useful to allow both Stop and remove queue in a single call.
> *Repro:*
> {code:java}
> Capacity-Scheduler.xml:
> yarn.scheduler.capacity.root.queues = new, default, dummy
> yarn.scheduler.capacity.root.default.capacity = 60
> yarn.scheduler.capacity.root.dummy.capacity = 30
> yarn.scheduler.capacity.root.new.capacity = 10
> curl -v -X PUT -d @abc.xml -H "Content-type: application/xml" 'http://<RM_IP>:8088/ws/v1/cluster/scheduler-conf'
> abc.xml
> <sched-conf>
> <update-queue>
> <queue-name>root.default</queue-name>
> <params>
> <entry>
> <key>capacity</key>
> <value>70</value>
> </entry>
> </params>
> </update-queue>
> <update-queue>
> <queue-name>root.new</queue-name>
> <params>
> <entry>
> <key>state</key>
> <value>STOPPED</value>
> </entry>
> </params>
> </update-queue>
> <remove-queue>root.new</remove-queue>
> </sched-conf>
> {code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: yarn-issues-help@hadoop.apache.org