You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hive.apache.org by "Shannon Ladymon (JIRA)" <ji...@apache.org> on 2016/03/29 00:51:25 UTC

[jira] [Commented] (HIVE-12290) Native Vector ReduceSink

    [ https://issues.apache.org/jira/browse/HIVE-12290?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15215034#comment-15215034 ] 

Shannon Ladymon commented on HIVE-12290:
----------------------------------------

Doc done:
* [Configuration Properties - hive.vectorized.execution.reducesink.new.enabled | https://cwiki.apache.org/confluence/display/Hive/Configuration+Properties#ConfigurationProperties-hive.vectorized.execution.reducesink.new.enabled]

> Native Vector ReduceSink
> ------------------------
>
>                 Key: HIVE-12290
>                 URL: https://issues.apache.org/jira/browse/HIVE-12290
>             Project: Hive
>          Issue Type: Improvement
>          Components: Hive
>            Reporter: Matt McCline
>            Assignee: Matt McCline
>            Priority: Critical
>             Fix For: 2.0.0
>
>         Attachments: HIVE-12290.01.patch, HIVE-12290.02.patch, HIVE-12290.03.patch, HIVE-12290.04.patch, HIVE-12290.05.patch, HIVE-12290.06.patch
>
>
> Currently, VectorReduceSinkOperator is a pass-thru to ReduceSinkOperator, so we incur object inspector costs.
> Native vectorization will not use object inspectors and allocate memory up front that will be reused for each batch.



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