[jira] [Commented] (OOZIE-2885) Running Spark actions should not need Hive on the classpath

["Satish Subhashrao Saley (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/OOZIE-2885?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16172503#comment-16172503 ] 

Satish Subhashrao Saley commented on OOZIE-2885:
------------------------------------------------

OOZIE-2845 removed reflection based code along with the try-catch block [click here for diff |https://github.com/apache/oozie/commit/6bac84745b9c62907e8cc6a16bad6c76ac3eb9c6#diff-f6e9d67b7671f1c2eb168642bd79789aL596].
OOZIE-2799 added some properties to hive-site.xml for users accessing hive via spark.
I think we should put try-catch block back rather than including hive dependencies in spark sharelib. What do you think [~gezapeti]?

> Running Spark actions should not need Hive on the classpath
> -----------------------------------------------------------
>
>                 Key: OOZIE-2885
>                 URL: https://issues.apache.org/jira/browse/OOZIE-2885
>             Project: Oozie
>          Issue Type: Bug
>            Reporter: Peter Cseh
>
> SparkMain references HiveConf so we need hive-common jar on the classpath even if the Spark job has nothing to do with Hive. 
> This shouldn't be the case.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)