feat(rbac): grant inference-operator access to InferenceGatewayConfig - #463
Merged
Merged
Conversation
chad119
approved these changes
Sep 16, 2026
Sean1783
approved these changes
Sep 16, 2026
papriwal
approved these changes
Sep 16, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Grant the inference-operator ClusterRole access to
InferenceGatewayConfig:inferencegatewayconfigs: full CRUD (get;list;watch;create;update;patch;delete) — the operator upserts a scheduler entry on the shared gateway CR when a model opts in.inferencegatewayconfigs/status:getonly — the operator mirrors gateway state onto the model's status; the SageMaker Inference Gateway controller owns status writes.No
/finalizersverb: the operator never finalizes the IGC (the gateway owns its lifecycle).Why
Follow-up to #461, which added the
spec/status.inferenceGatewayopt-in fields to the JumpStartModel / InferenceEndpointConfig CRDs and bumped the operator to v3.6, but did not grant RBAC for the InferenceGatewayConfig resource.When a model sets
spec.inferenceGateway.enabled: true, the operator's first Get/Create/Update on the clusterInferenceGatewayConfigis rejected withforbiddenwithout these verbs, so operator↔gateway wiring never completes. The operator interacts with the IGC via unstructured (GVK + field paths), so no cross-repo Go/type dependency is introduced — RBAC is the only thing needed at the Kubernetes API layer.Scope
RBAC only; no chart/appVersion/image bump. The IGC CRD itself is installed by the
sagemaker-inference-gatewaychart.Notes
Mirrors the internal operator ClusterRole (kubebuilder markers) that already carries these rules. Minimal-privilege split (status
getonly) is intentional.