sqoop-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "David Robson (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (SQOOP-1388) One of Export scenario fails with an error : java.sql.SQLException: Unable to perform an "exchange subpartition" operation for the table ...
Date Mon, 21 Jul 2014 04:21:38 GMT

    [ https://issues.apache.org/jira/browse/SQOOP-1388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14068182#comment-14068182
] 

David Robson commented on SQOOP-1388:
-------------------------------------

This is caused because the requested number of splits is 4, but CombineFileInputFormat is
generating 5 splits. OraOop is assuming the requested number of splits will be the actual
number of splits. Need to confirm if this is meant to be the case or not.

> One of Export scenario fails with an error  : java.sql.SQLException: Unable to perform
an "exchange subpartition" operation for the table ...
> ---------------------------------------------------------------------------------------------------------------------------------------------
>
>                 Key: SQOOP-1388
>                 URL: https://issues.apache.org/jira/browse/SQOOP-1388
>             Project: Sqoop
>          Issue Type: Bug
>         Environment:  " sqoop-1.4.5-SNAPSHOT.bin__hadoop-2.0.4-alpha.tar.gz " on CDH4
(melhadoop 01).
>            Reporter: Vidya Iyengar
>            Assignee: David Robson
>
> Ran Export Scenario 13 from the Guy Ley Mar document on SASDB01.
> These are the steps I followed for CASE 13.
>   
> drop table junk;
> drop table junk13;
> create table junk as select * from dba_objects;
> sqoop import --direct --connect jdbc:oracle:thin:@melsasdb01:1521:sasdb01 --username
oraoop --password oraoop --table junk  --where "OBJECT_ID<100" --target-dir junk_a
> sqoop import --direct --connect jdbc:oracle:thin:@melsasdb01:1521:sasdb01 --username
oraoop --password oraoop --table junk  --where "OBJECT_ID>= 100 and OBJECT_ID < 200"
--target-dir junk_b
> sqoop export -D oraoop.drop.table=true -Doraoop.template.table=oraoop.junk -Doraoop.nologging=true
-Doraoop.partitioned=true -Doraoop.export.merge=false --direct --connect jdbc:oracle:thin:@melsasdb01:1521:sasdb01
--username oraoop --password oraoop --table junk13 --export-dir junk_a
> sqoop export  -Doraoop.export.merge=false --direct --connect jdbc:oracle:thin:@melsasdb01:1521:sasdb01
--username oraoop --password oraoop --table junk13 --export-dir junk_b
> Note: Failed when i ran the first export.
> On the Hadoop file system, only one file contains data remaining 3 files are with size
0. 
> Removing empty files export worked ok.
> Imported folders are available junk_a_bug , junk_b_bug for debugging.
> Attached file contains the output error.



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

Mime
View raw message