hadoop-yarn-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "sandflee (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (YARN-5022) NM shutdown takes too much time
Date Mon, 02 May 2016 04:41:12 GMT

    [ https://issues.apache.org/jira/browse/YARN-5022?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15266110#comment-15266110
] 

sandflee commented on YARN-5022:
--------------------------------

this is mainly caused by NonAggregationLogHandler,

bq. 1, 4s for containerManager to wait all applications removed from context
applications couldn't be removed util log deleter is executed in NonAggregationLogHandler,
this is delayed 63hours by default
bq. 2, 10s for NonAggregatingLogHandler to wait delayed log deleter
NonAggregatingLogHandler#serviceStop() will wait all log deleter to be executed



> NM shutdown takes too much time
> -------------------------------
>
>                 Key: YARN-5022
>                 URL: https://issues.apache.org/jira/browse/YARN-5022
>             Project: Hadoop YARN
>          Issue Type: Improvement
>            Reporter: sandflee
>            Assignee: sandflee
>         Attachments: YARN-5022.01.patch, nm.log
>
>
> start NM with nm recovery disabled and log aggregation disabled.
> It takes about 14s for nm to quit while receiving SIGTERM, 
> 1, 4s for containerManager to wait all applications removed from context
> 2, 10s for NonAggregatingLogHandler to wait delayed log deleter



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: yarn-issues-help@hadoop.apache.org


Mime
View raw message