[jira] [Commented] (TEZ-3573) Allow user to cap number of task in cartesian product (unpartitioned case)

["Siddharth Seth (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/TEZ-3573?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16083114#comment-16083114 ] 

Siddharth Seth commented on TEZ-3573:
-------------------------------------

[~aplusplus] - is this still relevant after recent changes to CartesianProduct?

> Allow user to cap number of task in cartesian product (unpartitioned case)
> --------------------------------------------------------------------------
>
>                 Key: TEZ-3573
>                 URL: https://issues.apache.org/jira/browse/TEZ-3573
>             Project: Apache Tez
>          Issue Type: Sub-task
>            Reporter: Zhiyuan Yang
>            Assignee: Zhiyuan Yang
>         Attachments: TEZ-3573.1.patch, TEZ-3573.2.patch, TEZ-3573.3.patch
>
>
> Auto grouping can help reduce #tasks in cartesian product but may still result in too many tasks in case of huge input data. It will be useful for user to cap #task, so that cartesian product won't abuse available resource. The primary limiter will still be auto grouping, but this will be a hard limit which cannot be exceeded anyway.



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