[
https://issues.apache.org/jira/browse/HIVE-20503?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16606730#comment-16606730
]
Ashutosh Chauhan commented on HIVE-20503:
-----------------------------------------
Actually longer term we will abandon optimized hashtable altogether. We cant
just yet because fast doesnt support decimals yet. With 20491 we do estimates
using fast which is more conservative. So, I think its not worth the
complexity in compiler to account for fast vs optimized. I think we can stop at
HIVE-20491 and not pursue this any further. Fast will also be made more memory
efficient with HIVE-20513 .
> Use datastructure aware estimations during mapjoin selection
> ------------------------------------------------------------
>
> Key: HIVE-20503
> URL: https://issues.apache.org/jira/browse/HIVE-20503
> Project: Hive
> Issue Type: Bug
> Components: Statistics
> Reporter: Zoltan Haindrich
> Assignee: Zoltan Haindrich
> Priority: Major
> Attachments: HIVE-20503.01.patch, HIVE-20503.01wip01.patch,
> HIVE-20503.01wip01.patch, HIVE-20503.02.patch
>
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
