younsl opened a new pull request, #420: URL: https://github.com/apache/superset-kubernetes-operator/pull/420
## Summary Add an optional [priorityClassName](https://kubernetes.io/docs/concepts/scheduling-eviction/pod-priority-preemption/) value for the operator manager pod. Today the chart has no way to set a PriorityClass on its own Deployment, and extraManifests cannot patch it, so clusters that rank platform controllers above workloads cannot protect the operator from preemption or schedule it ahead of tenant pods. The value defaults to null, so the rendered manifest is unchanged and pods keep the cluster default priority. ## Details - The Deployment template renders priorityClassName only when the value is set, next to the other scheduling fields (nodeSelector, affinity, tolerations, topologySpreadConstraints). - The default follows the null convention for optional strings from #394. [values.schema.json](https://helm.sh/docs/topics/charts/#schema-files) accepts a string or null, and an explicit empty string still renders nothing. - The values description links the [Kubernetes pod priority docs](https://kubernetes.io/docs/concepts/scheduling-eviction/pod-priority-preemption/). The chart README is regenerated with [helm-docs](https://github.com/norwoodj/helm-docs). - The full-options test values set the key so the values coverage check passes, and the full-options snapshot gains the field. Two focused tests cover the default (field absent) and a configured value. - The changelog entry is added under Unreleased. <details> <summary>Verification</summary> - [helm-unittest](https://github.com/helm-unittest/helm-unittest): 31 tests and 26 snapshots pass. - helm lint strict and [kubeconform](https://github.com/yannh/kubeconform) strict pass for the default and full-options values (10 and 15 resources, all valid). - make helm-lint, make helm-values-covered, make lint-md, and make codegen (no diff) pass. - kind cluster (Kubernetes 1.37): with an existing PriorityClass, the manager pod is admitted with that class and its resolved priority value (100000). With a nonexistent class, the ReplicaSet fails pod creation with "no PriorityClass with name missing-class was found", as expected from the Priority admission plugin. </details> -- 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]
