[
https://issues.apache.org/jira/browse/OOZIE-3332?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16591216#comment-16591216
]
Hadoop QA commented on OOZIE-3332:
----------------------------------
PreCommit-OOZIE-Build started
> Update the spark action examples to make it easier to run them on yarn in
> client and cluster mode
> -------------------------------------------------------------------------------------------------
>
> Key: OOZIE-3332
> URL: https://issues.apache.org/jira/browse/OOZIE-3332
> Project: Oozie
> Issue Type: Improvement
> Components: examples
> Reporter: Daniel Becker
> Assignee: Daniel Becker
> Priority: Trivial
> Attachments: OOZIE-3332.2.patch, OOZIE-3332.patch
>
>
> The spark and pyspark examples do not have a <mode> element in their
> workflow.xml files, probably because of the old way the spark CLI worked, in
> which to run spark on yarn in client mode, you would write
> {code:java}
> --master yarn-client
> {code}
> on the command line
> (https://spark.apache.org/docs/1.5.0/running-on-yarn.html). This still works,
> but from spark 1.6, the spark documentation
> (https://spark.apache.org/docs/1.6.0/running-on-yarn.html) says the way to do
> it is
> {code:java}
> --master yarn --mode client
> {code}
> .
> We should update the two examples to have <mode> elements in workflow.xml and
> corresponding values in job.properties.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
