[jira] Updated: (HIVE-108) sampling does not use internal column name

["Carl Steinbach (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/HIVE-108?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Carl Steinbach updated HIVE-108:
--------------------------------

    Fix Version/s: 0.3.0

> sampling does not use internal column name 
> -------------------------------------------
>
>                 Key: HIVE-108
>                 URL: https://issues.apache.org/jira/browse/HIVE-108
>             Project: Hadoop Hive
>          Issue Type: Bug
>          Components: Query Processor
>            Reporter: Namit Jain
>            Assignee: Namit Jain
>             Fix For: 0.3.0
>
>         Attachments: patch.txt
>
>
> sampling does not use internal column name, it uses column table name instead - which will not work in case there is any operator between table
> scan and sampling, which is possible with column pruning.
> The following test breaks: 
> (assuming srcbucket has more than 1 column: in this case 2 columns: key and value)
> -- no input pruning, sample filter
> EXPLAIN
> SELECT s.key
> FROM srcbucket TABLESAMPLE (BUCKET 1 OUT OF 5 on key) s;
> SELECT s.key
> FROM srcbucket TABLESAMPLE (BUCKET 1 OUT OF 5 on key) s;

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