drill-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jacques Nadeau (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (DRILL-882) Join between hive table and parquet file fail
Date Thu, 10 Jul 2014 20:33:04 GMT

    [ https://issues.apache.org/jira/browse/DRILL-882?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14057914#comment-14057914
] 

Jacques Nadeau commented on DRILL-882:
--------------------------------------

Fixed the CanNotPlanExceptin with code change in Optiq. 

select p.p_partkey
   from hive.part p, cp.`tpch/partsupp.parquet` ps
   where p.p_partkey = cast(ps.ps_partkey as int)
order by p.p_partkey
limit 20;  

+------------+
| p_partkey  |
+------------+
| 1          |
| 1          |
| 1          |
| 1          |
| 2          |
| 2          |
| 2          |
| 2          |
| 3          |
| 3          |
| 3          |
| 3          |
| 4          |
| 4          |
| 4          |
| 4          |
| 5          |
| 5          |
| 5          |
| 5          |
+------------+
20 rows selected (0.625 seconds)
 
Note, if we remove the "cast" function, Drill will return 0 row. That's a different issue,
which will be addressed seperately.

> Join between hive table and parquet file fail
> ---------------------------------------------
>
>                 Key: DRILL-882
>                 URL: https://issues.apache.org/jira/browse/DRILL-882
>             Project: Apache Drill
>          Issue Type: Bug
>          Components: Query Planning & Optimization
>            Reporter: Ramana Inukonda Nagaraj
>            Assignee: Jinfeng Ni
>            Priority: Critical
>             Fix For: 1.0.0-BETA1
>
>
> The following query fails with a cannot plan error
> select p.p_partkey 
>    from hive.part p, `tpch-multi/partsupp` ps 
>    where p.p_partkey = ps.ps_partkey 
>               and p.p_size = 41  
> order by p.p_partkey
> limit 20;
> The below queries work fine implying nothing is wrong with the source
> select p.p_partkey 
>    from hive.part p;
>    
> select ps.ps_partkey from `tpch-multi/partsupp` ps;
> The same query also works when both sides of join is from parquet or hive. Its only when
they are different that I get the below cannot plan error. 
> message: "Failure while parsing sql. < CannotPlanException:[ Node [rel#2666:Subset#26.PHYSICAL.SINGLETON([]).[]]
could not be implemented; planner state:



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Mime
View raw message