You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@tez.apache.org by "Yi Zhang (Jira)" <ji...@apache.org> on 2023/09/06 00:55:00 UTC

[jira] [Commented] (TEZ-4513) Add feature to fail DAG when too many re-runs

    [ https://issues.apache.org/jira/browse/TEZ-4513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17762238#comment-17762238 ] 

Yi Zhang commented on TEZ-4513:
-------------------------------

[~rbalamohan] wonder if there is some config to achieve same effect without a new config? thank you!

> Add feature to fail DAG when too many re-runs
> ---------------------------------------------
>
>                 Key: TEZ-4513
>                 URL: https://issues.apache.org/jira/browse/TEZ-4513
>             Project: Apache Tez
>          Issue Type: Improvement
>    Affects Versions: 0.10.2
>            Reporter: Yi Zhang
>            Priority: Major
>
> Sometimes when nodes failure happen, shuffle data are lost and producer tasks are re-run, those tasks' ancestor in turn may need to re-run, but cluster may not have enough resources to re-run those tasks fast. In this scenario, it may be desirable to fail the DAG.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)