[YARN-6483] Add nodes transitioning to DECOMMISSIONING state to the list of updated nodes returned by the Resource Manager as a response to the Application Master heartbeat

[Juan Rodríguez Hortalá]

Hi,

I have submitted a pull request https://github.com/apache/hadoop/pull/289
 for https://issues.apache.org/jira/browse/YARN-6483. This change is an
alternative proposal to YARN-3224, for notifying application masters about
the decommission of a node, that could help completing
https://issues.apache.org/jira/browse/YARN-914. This could be useful for
implementing functionality like for example
https://github.com/apache/spark/pull/19267, where a Spark AM is able to
blacklist executors when their corresponding node transitions into
DECOMMISSIONING state.

I would like to hear your feedback on this.

Thanks,

Juan Rodriguez Hortala