zheniantoushipashi commented on PR #37802: URL: https://github.com/apache/spark/pull/37802#issuecomment-1238949312
I studied your previous discussions, but i still think the change i pull request is useful， i investigative this document https://docs.google.com/document/d/1xgQGRpaHQX6-QH_J9YV2C2Dh6RpXefUpLM7KGkzL6Fg/edit#heading=h.1quyr1r2kr5n , i known spark will support more scheduler in the future, but the configuration i add in this pr is just for volcano scheduler , and the change is just in the VolcanoFeatureStep.scala , and the configuration prefixed with spark.kubernetes.scheduler.volcano.podGroup , not a global spark configuration for all the scheduler, We can arbitrarily create queues in volcano, and In our scenario, we need to dynamically specify the volcano queue， our users can dynamically use different queues , if we must create a template file Just to dynamically use the queue I created， it is very inconvenient， i must add some code to create a file or i must use configmap to mount the a new template file and restart the k8s application, it is no cool, i think the template file and the configuration is not conflict，  spark driver pod and executor pod can use template define the pod resource, and spark configuration can override the template file, I think this usage is reasonable。 -- 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]
