ExternalGroup CRD isn't being deployed (Helm)

[jlazerus]

What is the issue?

In version 2.16.1 of the linkerd-crds Helm chart, the external-group.yaml file is in a location that Helm does not process, resulting in deployment issues.

Problem Details:

• The external-group.yaml file is currently located at:
  root -> templates -> workload -> external-group.yaml
  which is not processed by Helm.
• In contrast, the external-workload.yaml file, which deploys correctly, is located within a subdirectory of charts/linkerd-crds:
  root -> charts/linkerd-crds -> templates -> workload -> external-workload.yaml.
  This placement allows it to be correctly processed by Helm.

Directory Structure Example:

Below is a simplified view of the current directory structure highlighting the problematic location of external-group.yaml:

├── Chart.lock
├── Chart.yaml
├── README.md
├── charts
│   └── linkerd-crds
│       ├── templates
│       │   └── workload
│       │       └── external-workload.yaml
└── templates
    └── workload
        └── external-group.yaml   <-- Not processed by Helm

How can it be reproduced?

1. Your values.yaml contains the line:
   manageExternalWorkloads: true
2. Deploy the linkerd-crds Helm Chart version 2.16.1 with 'helm install...`
3. Check for external-group.yaml Deployment:
   After the chart is installed, check the Kubernetes resources to see if the external-group.yaml CRD has been created:
   kubectl get crd | grep "external-group"
4. Verify the Missing Resource:
   You should see that the external-group resource is not created, indicating that the file in templates/workload/external-group.yaml was not processed by Helm.

Expected Outcome
The external-group.yaml CRD should be deployed as part of the Helm chart, similar to how the external-workload.yaml CRD is deployed.
Actual Outcome
The external-group.yaml CRD is missing from the Kubernetes cluster, as the file location prevents it from being processed by Helm.

Logs, error output, etc

output of linkerd check -o short

❯ linkerd check -o short
linkerd-cni-plugin
------------------
× cni plugin ConfigMap exists
    configmaps "linkerd-cni-config" not found
    see https://linkerd.io/2/checks/#cni-plugin-cm-exists for hints

| Running multicluster extension check
...hangs

Probably not relevant as everything else is running correctly in this cluster

Environment

Kubernetes current version: 1.30.5-gke.1014001

❯ linkerd version
Client version: enterprise-2.16.0
Server version: enterprise-2.16.0

Possible solution

Please consider relocating external-group.yaml to the charts/linkerd-crds/templates/workload/ directory, or to another directory that Helm processes, for consistency and correct deployment.

Additional context

No response

Would you like to work on fixing this bug?

no