hadoop-hive-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Namit Jain (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HIVE-969) choose the largest table for join as the table to be streamed
Date Thu, 03 Dec 2009 22:56:20 GMT

    [ https://issues.apache.org/jira/browse/HIVE-969?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12785600#action_12785600
] 

Namit Jain commented on HIVE-969:
---------------------------------

No, I was thinking of getting the size by directly making a hadoop call, or getting it from
the FileSystem API - 
we can change metastore once we have the complete plan

> choose the largest table for join as the table to be streamed
> -------------------------------------------------------------
>
>                 Key: HIVE-969
>                 URL: https://issues.apache.org/jira/browse/HIVE-969
>             Project: Hadoop Hive
>          Issue Type: Improvement
>          Components: Query Processor
>            Reporter: Namit Jain
>
> In absence of statistics, the join order is random. This creates a problem for the users
since mostly they dont care about the size of the tables
> So, instead of choosing the join order randomly, we can make the largest table as the
table to be streamed if the user did not specify any hints.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message