[
https://issues.apache.org/jira/browse/HIVE-11386?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14643505#comment-14643505
]
Matt McCline commented on HIVE-11386:
-------------------------------------
Phase 1 is restricted to aggregations involving LONGs (boolean, tinyint,
smallint, int, bigint/long) and aggregates that produce LONGs (min, max, sum,
count).
And, a *single* LONG key.
> Improve Vectorized GROUP BY Performance (Phase 1)
> -------------------------------------------------
>
> Key: HIVE-11386
> URL: https://issues.apache.org/jira/browse/HIVE-11386
> Project: Hive
> Issue Type: Bug
> Components: Hive
> Reporter: Matt McCline
> Assignee: Matt McCline
> Priority: Critical
> Attachments: HIVE-11386.01.patch
>
>
> Improve vectorized GROUP BY performance, with an eye towards the new LLAP
> memory management (dramatically reduce the number of Java object, allocate
> very large objects, etc).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
