Fix events RBAC and automate Helm ClusterRole generation from markers#292
Merged
Fix events RBAC and automate Helm ClusterRole generation from markers#292
Conversation
…y from Go with make generate
carlydf
commented
Apr 23, 2026
carlydf
commented
Apr 23, 2026
jaypipes
reviewed
Apr 23, 2026
Contributor
jaypipes
left a comment
jaypipes
left a comment
There was a problem hiding this comment.
lgtm.
Might I suggest adding a check to the .github/workflows/helm-validate.yml workflow file that double-checks nobody manually updated the Helm RBAC stuff?
jobs:
helm-check-rbac:
name: Check Helm RBAC wasn't manually updated
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- run: make manifests && git diff --exit-code helm/temporal-worker-controller/templates/rbac.yaml
shashwatsuri
pushed a commit
to shashwatsuri/temporal-worker-controller
that referenced
this pull request
Apr 28, 2026
…temporalio#292) ## Problem ### Two related issues: 1. Events permission used the wrong API group. The +kubebuilder:rbac marker for events specified groups=events.k8s.io (the newer structured events API), but controller-runtime records v1.Event objects against the core "" API group. This caused the controller to log Server rejected event (will not retry!) errors when deployed in a different namespace from the TemporalWorkerDeployment CRs it manages — common cluster-wide deployment pattern. 2. The Helm ClusterRole was hand-maintained and had drifted from the Go markers. The +kubebuilder:rbac markers in worker_controller.go were incomplete: workerresourcetemplates, workerresourcetemplates/status, and subjectaccessreviews were all present in the Helm chart but had no corresponding markers. This made it easy for RBAC rules to fall out of sync — which is exactly how the events bug happened in the first place. ## Changes Fix the events API group (worker_controller.go, config/rbac/role.yaml): correct groups=events.k8s.io → groups=core in the marker; the generated manifest now uses apiGroups: [""]. Add missing markers (worker_controller.go): add markers for workerresourcetemplates (get/list/watch/patch/update), workerresourcetemplates/status (get/patch/update), and authorization.k8s.io/subjectaccessreviews (create) to match what was already deployed by the Helm chart. Automate Helm ClusterRole sync (hack/sync-rbac-rules.py, Makefile, helm/.../rbac.yaml): make manifests now runs a script that reads the controller-gen-generated config/rbac/role.yaml and replaces the # GENERATED RULES BEGIN / # GENERATED RULES END section in the Helm template. The Helm-templated dynamic rules (the allowedResources range) are left untouched. Going forward, adding a +kubebuilder:rbac marker and running make manifests is all that's needed to update the deployed ClusterRole. Closes temporalio#277
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
2 participants
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.
Problem
Two related issues:
Events permission used the wrong API group.
The +kubebuilder:rbac marker for events specified groups=events.k8s.io (the newer structured events API), but controller-runtime records v1.Event objects against the core "" API group. This caused the controller to log Server rejected event (will not retry!) errors when deployed in a different namespace from the TemporalWorkerDeployment CRs it manages — common cluster-wide deployment pattern.
The Helm ClusterRole was hand-maintained and had drifted from the Go markers.
The +kubebuilder:rbac markers in worker_controller.go were incomplete: workerresourcetemplates, workerresourcetemplates/status, and subjectaccessreviews were all present in the Helm chart but had no corresponding markers. This made it easy for RBAC rules to fall out of sync — which is exactly how the events bug happened in the first place.
Changes
Fix the events API group (worker_controller.go, config/rbac/role.yaml): correct groups=events.k8s.io → groups=core in the marker; the generated manifest now uses apiGroups: [""].
Add missing markers (worker_controller.go): add markers for workerresourcetemplates (get/list/watch/patch/update), workerresourcetemplates/status (get/patch/update), and authorization.k8s.io/subjectaccessreviews (create) to match what was already deployed by the Helm chart.
Automate Helm ClusterRole sync (hack/sync-rbac-rules.py, Makefile, helm/.../rbac.yaml): make manifests now runs a script that reads the controller-gen-generated config/rbac/role.yaml and replaces the # GENERATED RULES BEGIN / # GENERATED RULES END section in the Helm template. The Helm-templated dynamic rules (the allowedResources range) are left untouched. Going forward, adding a +kubebuilder:rbac marker and running make manifests is all that's needed to update the deployed ClusterRole.
Closes #277