[
https://issues.apache.org/jira/browse/YARN-178?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13482432#comment-13482432
]
Robert Joseph Evans commented on YARN-178:
------------------------------------------
I am fine with either setting it through a contructor or through a setter
method. I agree with Radim that the constructor makes it more obvious that it
is not supposed to change, but other places tend to use setter methods instead,
like with Configuration which is also not supposed to change.
> Fix custom ProcessTree instance creation
> ----------------------------------------
>
> Key: YARN-178
> URL: https://issues.apache.org/jira/browse/YARN-178
> Project: Hadoop YARN
> Issue Type: Bug
> Affects Versions: 3.0.0, 0.23.5
> Reporter: Radim Kolar
> Assignee: Radim Kolar
> Priority: Critical
> Attachments: pstree-instance2.txt, pstree-instance.txt
>
>
> 1. In current pluggable resourcecalculatorprocesstree is not passed root
> process id to custom implementation making it unusable.
> 2. pstree do not extend Configured as it should
> Added constructor with pid argument with testsuite. Also added test that
> pstree is correctly configured.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
