You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hive.apache.org by "Gopal V (JIRA)" <ji...@apache.org> on 2017/06/07 01:32:18 UTC

[jira] [Updated] (HIVE-12492) MapJoin: 4 million unique integers seems to be a probe plateau

     [ https://issues.apache.org/jira/browse/HIVE-12492?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Gopal V updated HIVE-12492:
---------------------------
    Description: 
After 4 million keys, the map-join implementation seems to suffer from a performance degradation. 

The hashtable build & probe time makes this very inefficient, even if the data is very compact (i.e 2 ints).

Falling back onto the shuffle join or bucket map-join is useful after 2^22 items.

(Note: this fixes a statsutil issue - due to the extra clone() in the column stats path)

  was:
After 4 million keys, the map-join implementation seems to suffer from a performance degradation. 

The hashtable build & probe time makes this very inefficient, even if the data is very compact (i.e 2 ints).

Falling back onto the shuffle join or bucket map-join is useful after 2^22 items.


> MapJoin: 4 million unique integers seems to be a probe plateau
> --------------------------------------------------------------
>
>                 Key: HIVE-12492
>                 URL: https://issues.apache.org/jira/browse/HIVE-12492
>             Project: Hive
>          Issue Type: Improvement
>          Components: Query Planning
>    Affects Versions: 1.3.0, 1.2.1, 2.0.0
>            Reporter: Gopal V
>            Assignee: Jesus Camacho Rodriguez
>              Labels: TODOC2.2
>             Fix For: 2.2.0
>
>         Attachments: HIVE-12492.01.patch, HIVE-12492.02.patch, HIVE-12492.patch
>
>
> After 4 million keys, the map-join implementation seems to suffer from a performance degradation. 
> The hashtable build & probe time makes this very inefficient, even if the data is very compact (i.e 2 ints).
> Falling back onto the shuffle join or bucket map-join is useful after 2^22 items.
> (Note: this fixes a statsutil issue - due to the extra clone() in the column stats path)



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)