[
https://issues.apache.org/jira/browse/SQOOP-2757?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dian Fu updated SQOOP-2757:
---------------------------
Attachment: SQOOP-2757.001.patch
> Sqoop2: Add module connector-sdk-hadoop to hold hadoop specific SDK classes
> used by the connectors
> --------------------------------------------------------------------------------------------------
>
> Key: SQOOP-2757
> URL: https://issues.apache.org/jira/browse/SQOOP-2757
> Project: Sqoop
> Issue Type: Sub-task
> Reporter: Dian Fu
> Assignee: Dian Fu
> Fix For: 1.99.7
>
> Attachments: SQOOP-2757.001.patch
>
>
> As discussed in the review board of SQOOP-2752:
> {quote}
> I would prefer not to put hadoop as a depedency on connector-sdk. This would
> force every connector to synchronize with Hadoop dependencies (guava and
> such). Can we instead create a new module "connector-sdk-hadoop" that will
> contain Hadoop specific SDK classes and will in turn contain this Hadoop
> dependency (arguably if the connector needs something from Hadoop they should
> be fine with having hadoop jars on classpath). This way the connectors that
> doesn't need anything from hadoop don't end up having Hadoop's guava (and
> others) on the classpath.
> {quote}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
