Pluies opened a new issue, #779: URL: https://github.com/apache/spark-kubernetes-operator/issues/779
Hello, First, thank you very much for the work on the Spark Kubernetes Operator, it's a breath of fresh air and makes setting up Spark clusters very easy and straightforward 🙏 I wanted to enquire about plans to support upgrades based on the CRDs. 👀 For example, let's say we are running the `pi` SparkApp from [`examples/pi.yaml`](https://github.com/apache/spark-kubernetes-operator/blob/main/examples/pi.yaml), and we edit the SparkApp to upgrade the Spark version from `4.2.0` to `4.3.0`. Currently, this is a noop: the resource will accept the change, but the operator will silently ignore it, and will not reconcile anything. The Spark app itself will keep running on Spark 4.2.0. The only workaround at the moment is to `kubectl delete` that resource, then recreate it with an updated `spec`. It's a bit cumbersome, and doesn't work well with standard ways of deploying manifests to k8s (gitops, helm update, even kubectl apply). Usually, operators detect changes in spec and reconcile the running objects underneath so that they match the desired specs. I've searched the Github issues in this repository, as well as JIRA issues for the Kubernetes Operator, but couldn't find relevant tickets. I'd be happy to help contribute towards this feature, but first I wanted to check with the maintainers: - whether you agree this should be within the remit of this operator - if you had any specific concerns or worries about the implementation of this feature (And also check if maybe work has already started on this somewhere so we don't duplicate effort!) Thanks! - Florent cc @dongjoon-hyun 🙏 -- 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]
