[
https://issues.apache.org/jira/browse/FLINK-19828?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17221492#comment-17221492
]
Robert Metzger commented on FLINK-19828:
----------------------------------------
I see two approaches:
a) consider all yaml files in a specified directory Flink configurations
(instead of the current approach of looking for a specific file in the
specified directory)
b) introduce a new configuration parameter to point to an additional
configuration file
I prefer approach a). I believe we could sort the files by name, in case users
want to introduce overwrites. It seems that one can mount a config map to a
specific file in a directory:
https://kubernetes.io/docs/tasks/configure-pod-container/configure-pod-configmap/#add-configmap-data-to-a-specific-path-in-the-volume
> Allow loading multiple configuration files
> ------------------------------------------
>
> Key: FLINK-19828
> URL: https://issues.apache.org/jira/browse/FLINK-19828
> Project: Flink
> Issue Type: New Feature
> Components: Runtime / Configuration
> Affects Versions: 1.12.0
> Reporter: Robert Metzger
> Priority: Major
>
> This is a feature request from the mailing list:
> https://lists.apache.org/thread.html/r84ad3af6585ba8138d8bfa8d3d391dab9017aace52df8a757989be1e%40%3Cuser.flink.apache.org%3E
> For scenarios where you want to deploy Flink (for example on Kubernetes) with
> a set of common configuration keys (for example for the file system), and
> some deployment specific configuration keys (for example savepoint location),
> it would be helpful if Flink had a mechanism for loading multiple
> configuration files.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
