[
https://issues.apache.org/jira/browse/MAPREDUCE-907?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12755183#action_12755183
]
Aaron Kimball commented on MAPREDUCE-907:
-----------------------------------------
1) \-\-split-by specifies a column on which to split the current table. If it's
importing all tables, then any table-specific settings are meaningless.
\-\-all-tables does in fact use JDBC metadata to infer the primary key. (As
will \-\-table if \-\-split-by is not specified.)
2) Could do that.
3) The concepts seemed logically separate enough in my mind that they deserved
separate variables
4) Not necessarily. I can see this run() method being overridden and then
called with super.run(). Wrapping with class-by-conf seems like a bad idea,
this is exactly what inheritance is for.
5) +1 to the idea, though I feel like that's a separate API patch.
6) I wholeheartedly agree. There's definitely another JIRA for "More methods
that mapreduce.Job needs", but I think that's separate work.
> Sqoop should use more intelligent splits
> ----------------------------------------
>
> Key: MAPREDUCE-907
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-907
> Project: Hadoop Map/Reduce
> Issue Type: Improvement
> Components: contrib/sqoop
> Reporter: Aaron Kimball
> Assignee: Aaron Kimball
> Attachments: MAPREDUCE-907.patch
>
>
> Sqoop should use the new split generation / InputFormat in MAPREDUCE-885
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
