ssirowa14 commented on code in PR #925:
URL:
https://github.com/apache/spark-kubernetes-operator/pull/925#discussion_r4251673225
##########
spark-operator/src/main/java/org/apache/spark/k8s/operator/decorators/DriverResourceDecorator.java:
##########
@@ -69,13 +95,29 @@ public <T extends HasMetadata> T decorate(T resource) {
}
if (!ownerReferenceExists) {
log.debug("Adding OwnerReference to driver for secondary resource");
- ObjectMeta metaData =
+ Map<String, String> resourceLabels = resource.getMetadata().getLabels();
+ ObjectMetaBuilder builder =
new ObjectMetaBuilder(resource.getMetadata())
- .addToOwnerReferences(buildOwnerReferenceTo(driverPod))
- .addToLabels(driverPod.getMetadata().getLabels())
- .build();
- resource.setMetadata(metaData);
+ .addToOwnerReferences(buildOwnerReferenceTo(driverPod));
+ driverPod.getMetadata().getLabels().forEach((key, value) -> {
Review Comment:
> Have you considered propagating only the identity labels
(isManagedLabel(key)) to fix the root cause?
This actually would be a large behavior change and we may need more
scrutinized assessment.
Current behaviour of maintaining a label exclusion list and reapply all
labels that a resource explicitly defines (excluding the ones mentioned in
label exclusion list) is a very pin pointed and small surface area change.
We can explore the possibility of "propagating only the identity labels
during resource decoration" in a separate SPARK jira.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
