hadoop-hive-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ning Zhang (JIRA)" <j...@apache.org>
Subject [jira] Updated: (HIVE-1037) SerDe & ObjectInspector for RowContainer mismatch with the input data
Date Fri, 08 Jan 2010 20:43:54 GMT

     [ https://issues.apache.org/jira/browse/HIVE-1037?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel

Ning Zhang updated HIVE-1037:

    Attachment: HIVE-1037.patch

Uploading HIVE-1037.patch. Changes include:
1) moving initTableDesc from joinDesc to CommonJoinOperator to make it after column pruner.

2) some misc changes in RowContainer including better error reporting.
3) add a new unit test in join40.q

> SerDe & ObjectInspector for RowContainer mismatch with the input data
> ---------------------------------------------------------------------
>                 Key: HIVE-1037
>                 URL: https://issues.apache.org/jira/browse/HIVE-1037
>             Project: Hadoop Hive
>          Issue Type: Bug
>            Reporter: Ning Zhang
>            Assignee: Ning Zhang
>            Priority: Blocker
>             Fix For: 0.5.0
>         Attachments: HIVE-1037.patch
> In CommonJoinOperator, RowContainer is created for each input table with the SerDe and
ObjectInspector to serialize/deserialize that row to persistent storage. The serde/OI could
be null in the case of the value columns are pruned by column pruner. An example query is
> select count(1) from A join B on A.key=B.key;
> Another case of mismatch is that the tableDesc was initialized at compile time before
the column pruner take place. This could cause inconsistency in the SerDe/OI with the input
data. This should be moved to execution time when the join operator is initialized. 

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

View raw message