[
https://issues.apache.org/jira/browse/HADOOP-13717?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15629744#comment-15629744
]
Anu Engineer commented on HADOOP-13717:
---------------------------------------
[~andrew.wang] [~aw] Sorry to comment on this late. Just saw it when I was
looking at the git log. The DiskBalancerCLI runs but for a brief moment when
planning and writes a plan file out. Then when you submit plan it reads the
plan file and sends it datanode via RPC. Both of those AFAIK does not qualify
as a daemon behaviour. So why classify this as a daemon ?
I agree with the balancer comments, it does run for a long time. All long
running activity for diskbalancer is confined within the datanode, Perhaps I am
missing something here ?
> Normalize daemonization behavior of the diskbalancer with balancer and mover
> ----------------------------------------------------------------------------
>
> Key: HADOOP-13717
> URL: https://issues.apache.org/jira/browse/HADOOP-13717
> Project: Hadoop Common
> Issue Type: Bug
> Components: scripts
> Affects Versions: 3.0.0-alpha1
> Reporter: Andrew Wang
> Assignee: Andrew Wang
> Fix For: 3.0.0-alpha2
>
> Attachments: HADOOP-13717.001.patch
>
>
> Issue found when working with the HDFS balancer.
> In {{hadoop_daemon_handler}}, it calls {{hadoop_verify_logdir}} even for the
> "default" case which calls {{hadoop_start_daemon}}. {{daemon_outfile}} which
> specifies the log location isn't even used here, since the command is being
> started in the foreground.
> I think we can push the {{hadoop_verify_logdir}} call down into
> {{hadoop_start_daemon_wrapper}} instead, which does use the outfile.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
