[Bug] Helm Chart can't be rendered (v5.4.2)

[siegenthalerroger]

Describe the bug

When rendering the helm chart with FluxCD's helm-controller the chart fails to unmarshal certain values due to duplicates.

Helm install failed: error while running post render on files: map[string]interface {}(nil): yaml: unmarshal errors: line 16: mapping key "helm.sh/chart" already defined at line 7 line 14: mapping key "app.kubernetes.io/name" already defined at line 8 line 12: mapping key "app.kubernetes.io/instance" already defined at line 9 line 15: mapping key "app.kubernetes.io/version" already defined at line 10 line 13: mapping key "app.kubernetes.io/managed-by" already defined at line 11 line 32: mapping key "app.kubernetes.io/name" already defined at line 28 line 30: mapping key "app.kubernetes.io/instance" already defined at line 29

Version

v5.4.2

To Reproduce

Use FluxCD to reconcile a HelmRelease with the following values.

    # See: https://grafana-operator.github.io/grafana-operator/docs/installation/helm/#values
    namespaceScope: false # default=false
    # watchNamespaces: # default=* (all namespaces)  
    
    priorityClassName: cluster-medium

    resources:
      requests:
        cpu: 50m
        memory: 100Mi
      limits:
        cpu: 200m
        memory: 250Mi

Expected behavior

The helm chart to render and Flux to reconcile the resources as functional in v5.4.1.

Suspect component/Location where the bug might be occurring

The change that causes this was probably introduced by this PR: 6549cb2

Runtime (please complete the following information):

• OS: Azure Linux
• Grafana Operator Version: v5.4.2
• Environment: Kubernetes 1.26
• Deployment type: FluxCD HelmRelease

[weisdd]

I see that there are some duplicated labels in Deployment resource. It's not something that has recently been changed in the chart, it's more likely to be an error that one of the flux controllers stopped tolerating. Though, we can certainly fix that on the chart side as well.

[weisdd]

Surprising enough, you're right, that issue came from the rbac template. I guess the context got corrupted. Thanks for pointing at that! :)