hadoop-yarn-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jason Lowe (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (YARN-6774) YARN-site to have control over start sequence number for TASK Attempt IDs instead of hardcoded sequence starting at 1000s.
Date Fri, 07 Jul 2017 13:55:00 GMT

    [ https://issues.apache.org/jira/browse/YARN-6774?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16078106#comment-16078106
] 

Jason Lowe commented on YARN-6774:
----------------------------------

Task IDs are not assigned by YARN, they are assigned by a particular application framework
(e.g.: MapReduce, Tez).  Unlike MapReduce task attempt IDs, YARN container IDs have the application
attempt encoded within them, so their sequence number simply restarts at zero.  Are you really
referring to the logic for handling task IDs at https://github.com/apache/hadoop/blob/branch-2.7.1/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/job/impl/TaskImpl.java#L328-L332
?  If so then this is a change in MapReduce and mapred-site rather than YARN.

> YARN-site to have control over start sequence number for TASK Attempt IDs instead of
hardcoded sequence starting at 1000s. 
> ---------------------------------------------------------------------------------------------------------------------------
>
>                 Key: YARN-6774
>                 URL: https://issues.apache.org/jira/browse/YARN-6774
>             Project: Hadoop YARN
>          Issue Type: Wish
>          Components: yarn
>         Environment: OEL 6
>            Reporter: Deepak Chander
>            Priority: Minor
>
> When container attempts run for first application master attempt , it starts from 0.
> When application master fails, it starts from 1000. 
> The  logic of container id generation in below code.
> https://hadoop.apache.org/docs/r2.7.3/api/src-html/org/apache/hadoop/yarn/api/records/ContainerId.html#line.86
> - It would be nice if YARN-site to have control over start sequence number for TASK Attempt
IDs instead of hard coded sequence starting at 1000s. 



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

---------------------------------------------------------------------
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