Github user jacek-lewandowski commented on the pull request:
https://github.com/apache/spark/pull/3571#issuecomment-69045861
@vanzin I read about Yarn and Mesos deployments and found that making those
namespaces does not make much sense because you cannot have different
configurations in a single actor system.
For standalone deployment, it will work as I said before - the user can
provide his own configuration in SparkConf, but to fetch it, the worker
configuration will be used and then it will be overridden by the user
configuration.
For YARN, we still need some settings to fetch the configuration from the
driver. What I did was to repeat the same solution as it is applied for
spark.auth namespace - SSL settings are required before the actual
configuration is applied (the same for secret token - as it is implemented
right now).
What do you think? I really want to move this PR forward.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
