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 "Junping Du (JIRA)" <ji...@apache.org> on 2016/01/17 23:15:39 UTC

[jira] [Commented] (YARN-4060) Revisit default retry config for connection with RM

    [ https://issues.apache.org/jira/browse/YARN-4060?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15103947#comment-15103947 ] 

Junping Du commented on YARN-4060:
----------------------------------

Hi [~jianhe], do we have plan to deliver a fix in short term? If not, let's move it out of 2.6.4.

> Revisit default retry config for connection with RM 
> ----------------------------------------------------
>
>                 Key: YARN-4060
>                 URL: https://issues.apache.org/jira/browse/YARN-4060
>             Project: Hadoop YARN
>          Issue Type: Bug
>            Reporter: Jian He
>            Assignee: Jian He
>
> 15 minutes timeout for AM/NM connection with RM in non-ha scenario turns out to be  short in production environment.  The suggestion is to increase that to 30 min. Also, the retry-interval is set to 30 seconds which appears too long. We may reduce that to 10 seconds ?



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