[
https://issues.apache.org/jira/browse/HIVE-21399?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16786359#comment-16786359
]
Gopal V commented on HIVE-21399:
--------------------------------
That might explain why it isn't working for me right now - the config is very
optimistic and the stats say be pessimistic.
My config is 0.99 and I'm testing to see if a massive nDV product reduces it
and skips map aggregation faster.
> Adjust hive.map.aggr.hash.min.reduction statically depending on group by
> statistics
> -----------------------------------------------------------------------------------
>
> Key: HIVE-21399
> URL: https://issues.apache.org/jira/browse/HIVE-21399
> Project: Hive
> Issue Type: Improvement
> Components: Hive
> Reporter: Jesus Camacho Rodriguez
> Assignee: Jesus Camacho Rodriguez
> Priority: Major
> Attachments: HIVE-21399.01.patch, HIVE-21399.02.patch
>
>
> Currently, the value is set statically from config variable. If stats are
> available, we could try to adjust this value at optimization time to favor
> turning off hash aggregation earlier.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
