Yikun edited a comment on pull request #35359: URL: https://github.com/apache/spark/pull/35359#issuecomment-1024814048
@dongjoon-hyun Thanks ping me. > I don't think that the verbose revert commit and redo commit is necessary. We only need to revert commit, **redo commit is NOT needed**, First reason, we don't need the oringal PR (set configmap.metadata.namespace explictly), see validation in https://github.com/apache/spark/pull/35299#issuecomment-1019972817. And the original issue [SPARK-37713](https://issues.apache.org/jira/browse/SPARK-37713) should be marked as invalid I guess (might let @dcoliversun confirm it). Second reason, and also the [current namespace set way](https://github.com/apache/spark/blob/master/resource-managers/kubernetes/core/src/main/scala/org/apache/spark/deploy/k8s/submit/KubernetesClientUtils.scala#L93) break a special case like below(we shouldn't occupy any pontential data.key allowed by k8s in confFileMap): ``` apiVersion: v1 kind: ConfigMap metadata: name: configmap-test # This is metadata.namespace, the namespace of configMap namespace: non-default data: # This is user specified `namespace` file, fill up in configFileMap spark.kubernetes.namespace: foo=bar ``` So, I still think the revert might be needed due to above 2 reasons. -- 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]
