[jira] [Updated] (HIVE-20983) Vectorization: Scale up small hashtables, when collisions are detected

["Mustafa Iman (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/HIVE-20983?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Mustafa Iman updated HIVE-20983:
--------------------------------
    Attachment: HIVE-20983.2.patch

> Vectorization: Scale up small hashtables, when collisions are detected
> ----------------------------------------------------------------------
>
>                 Key: HIVE-20983
>                 URL: https://issues.apache.org/jira/browse/HIVE-20983
>             Project: Hive
>          Issue Type: Bug
>            Reporter: Gopal V
>            Assignee: Mustafa Iman
>            Priority: Major
>         Attachments: HIVE-20983.1.patch, HIVE-20983.2.patch
>
>
> Hive's hashtable estimates are getting better with HyperLogLog stats in place, but an accurate estimate does not always result in a low number of collisions.
> The hashtables which contain a very small number of items tend to lose their O(1) lookup performance where there are collisions. Since collisions are easy to detect within the fast hashtable implementation, a rehashing to a higher size will help these small hashtables avoid collisions and go back to O(1) perf.



--
This message was sent by Atlassian Jira
(v8.3.2#803003)