[
https://issues.apache.org/jira/browse/OOZIE-2687?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15693305#comment-15693305
]
Peter Cseh commented on OOZIE-2687:
-----------------------------------
We've discussed it with [~rkanter], [~pbacsko], [~asasvari] and [~andras.piros]:
We would like to introduce an optional tag in the xsd:
<launcher>
<memory>
// set memory for launcher container.
<vcores>// set core number for the launcher container.
<java-opts> //oozie.launcer.mapred.child.opts should be migrated here
.
Optional, one for each property.
<env> //… env
variables. Optional, one per each.
<queue> // yarn queue name for the launcher,
<sharelib> // one can define more that one sharelib here
</launcher>
This section could be set in workflow level or for each action separately.
Each property would have a default value in oozie-site.xml
Also, the schemas will be updated to have a <resourcemanager> tag instead of
the job-tracker. We'll leave the possibility to use job-tracker for backward
compatibility.
[~rohini], [~jaydeepvishwakarma], [~abhishekbafna], any ideas regarding this
change?
> OYA: Figure out how to handle MR-specific properties
> ----------------------------------------------------
>
> Key: OOZIE-2687
> URL: https://issues.apache.org/jira/browse/OOZIE-2687
> Project: Oozie
> Issue Type: Sub-task
> Reporter: Peter Cseh
>
> It was possible to manipulate the MapperLauncher's environment through
> properties like:
> # mapreduce.map.memory.mb
> # mapreduce.map.cpu.vcores
> # mapred.child.env
> # mapred.child.java.opts
> # mapred.job.queue.name - ability to set launcher queue
> E.g. We were using mapred.child.env to pass SPARK_HOME to the LauncherMapper
> and make PySpark work.
> Fixing OOZIE-2596 added a hack. We should decide how we support or break
> compatibility and how we allow the manipulation of the Launcher environment.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
