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

Posted to issues@flink.apache.org by "Chengxiang Li (JIRA)" <ji...@apache.org> on 2016/03/03 07:50:18 UTC

[jira] [Updated] (FLINK-3473) Add partial aggregate support in Flink

     [ https://issues.apache.org/jira/browse/FLINK-3473?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Chengxiang Li updated FLINK-3473:
---------------------------------
    Attachment: PartialAggregateinFlink_v2.pdf

> Add partial aggregate support in Flink
> --------------------------------------
>
>                 Key: FLINK-3473
>                 URL: https://issues.apache.org/jira/browse/FLINK-3473
>             Project: Flink
>          Issue Type: Improvement
>          Components: Table API
>            Reporter: Chengxiang Li
>            Assignee: Chengxiang Li
>         Attachments: PartialAggregateinFlink_v1.pdf, PartialAggregateinFlink_v2.pdf
>
>
> For decomposable aggregate function, partial aggregate is more efficient as it significantly reduce the network traffic during shuffle and parallelize part of the aggregate calculation. This is an umbrella task for partial aggregate.



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