You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hive.apache.org by "Edward Capriolo (JIRA)" <ji...@apache.org> on 2013/07/07 22:37:48 UTC

[jira] [Commented] (HIVE-4002) Fetch task aggregation for simple group by query

    [ https://issues.apache.org/jira/browse/HIVE-4002?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13701661#comment-13701661 ] 

Edward Capriolo commented on HIVE-4002:
---------------------------------------

This is a nice feature. There are times when I know that count(distinct(col)) and other operations like the one you have requested produce small result sets and the shuffle is the bottleneck. I do like this feature but turning it on manually is cumbersome for the end user.

I wonder if we can convert the last step at runtime somehow.(Probably not easily but that would be nice)
                
> Fetch task aggregation for simple group by query
> ------------------------------------------------
>
>                 Key: HIVE-4002
>                 URL: https://issues.apache.org/jira/browse/HIVE-4002
>             Project: Hive
>          Issue Type: Improvement
>          Components: Query Processor
>            Reporter: Navis
>            Assignee: Navis
>            Priority: Minor
>         Attachments: HIVE-4002.D8739.1.patch
>
>
> Aggregation queries with no group-by clause (for example, select count(*) from src) executes final aggregation in single reduce task. But it's too small even for single reducer because the most of UDAF generates just single row for map aggregation. If final fetch task can aggregate outputs from map tasks, shuffling time can be removed.
> This optimization transforms operator tree something like,
> TS-FIL-SEL-GBY1-RS-GBY2-SEL-FS + FETCH-TASK
> into 
> TS-FIL-SEL-GBY1-FS + FETCH-TASK(GBY2-SEL-LS)
> With the patch, time taken for auto_join_filters.q test reduced to 6 min (10 min, before).

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira