[jira] [Comment Edited] (TAJO-558) HCatalogStore can't scan columns.

["Jaehwa Jung (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/TAJO-558?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13884026#comment-13884026 ] 

Jaehwa Jung edited comment on TAJO-558 at 1/28/14 11:17 AM:
------------------------------------------------------------

I created a new RB https://reviews.apache.org/r/17453/.

Refactoring LogicalPlanner made this error, and I found that column qualifier just support pair of table name and column name. But if users use HCatalogStore, they can use different databases which stored on same HiveMetaStore. As a result, database name and table name and column name can be used at one column phase. So, I updated ColumnReferenceExpr and Schema to use database name.

You can verify it as follows:
{code:xml}
mvn clean install -Phcatalog-0.11.0
mvn clean install -Phcatalog-0.12.0
{code}

I already verified it on my local cluster.
Please review it. :)


was (Author: blrunner):
I created a new RB https://reviews.apache.org/r/17453/.

Refactoring LogicalPlanner made this error, and I found that column qualifier just support pair of table name and column name. But if users use HCatalogStore, they can use different databases which stored on same HiveMetaStore. As a result, database name and table name and column name can be used at one column phase. So, I updated ColumnReferenceExpr and Schema to use database name.


> HCatalogStore can't scan columns.
> ---------------------------------
>
>                 Key: TAJO-558
>                 URL: https://issues.apache.org/jira/browse/TAJO-558
>             Project: Tajo
>          Issue Type: Bug
>          Components: catalog
>    Affects Versions: 0.8-incubating
>            Reporter: Jaehwa Jung
>            Assignee: Jaehwa Jung
>             Fix For: 0.8-incubating
>
>
> If we write partitioned columns at where clause on tsql, tajo print a error message as follows:
> {code:xml}
> tajo> select * from table1 where id = 1;
> ERROR: no such a column name default
> {code}
> For reference, hive table schema is as the following:
> {code:xml}
> CREATE  TABLE table1(
>   id int, 
>   name string, 
>   score float, 
>   type string); 
> {code}
> It  occurred on both managed tables and partitioned tables.



--
This message was sent by Atlassian JIRA
(v6.1.5#6160)