hadoop-mapreduce-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "mingran wang (JIRA)" <j...@apache.org>
Subject [jira] Created: (MAPREDUCE-1449) Sqoop Documentation about --split-by column has to be unique key seems to be wrong
Date Wed, 03 Feb 2010 23:13:27 GMT
Sqoop Documentation about --split-by column has to be unique key seems to be wrong
----------------------------------------------------------------------------------

                 Key: MAPREDUCE-1449
                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-1449
             Project: Hadoop Map/Reduce
          Issue Type: Bug
          Components: contrib/sqoop
    Affects Versions: 0.20.1
            Reporter: mingran wang


http://archive.cloudera.com/docs/sqoo... 

The document above shows that " To guarantee correctness of your input, you must select an
ordering column for which each row has a unique value. If duplicate values appear in the ordering
column, the results of the import are undefined, and Sqoop will not be able to detect the
error." 

I read the source code for sqoop, it seems that the column to split by doesn't have to be
a unique key. Plus, when the primary key is a composite key, the sqoop code only takes the
first column of the composite key which in most cases is not unique key anyways. 

I also checked the output when non-unique key is used to split, there is nothing wrong with
the result. 

I am wondering if the document is wrong, or there is some hidden trickiness that I am not
aware of. 

I am using sqoop 20.1.

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


Mime
View raw message