Have A CI task to verify the CRD folder in sync

[jenting]

Describe the problem/challenge you have
[A description of the current limitation/problem/challenge that you are experiencing.]

Currently, the CI does not check the CRD folder YAML manifests are synced with the Velero image.

Describe the solution you'd like
[A clear and concise description of what you want to happen.]

Have a CI task to verify the CRD folder YAML manifests are synced with the Velero image.

Anything else you would like to add:
[Miscellaneous information that will assist in solving the issue.]

Environment:

• helm version (use helm version):
• helm chart version and app version (use helm list -n <YOUR NAMESPACE>):
• Kubernetes version (use kubectl version):
• Kubernetes installer & version:
• Cloud provider or hardware configuration:
• OS (e.g. from /etc/os-release):

[tungbq]

Hello, may I know the step verify the CRD folder YAML manifests are synced with the Velero image manually?

[jenting]

@tungbq
Compare the YAML between
velero install --crds-only --dry-run -o yaml and crds folder.

[tungbq]

@tungbq Compare the YAML between velero install --crds-only --dry-run -o yaml and crds folder.

Thanks @jenting for your prompt response.
I'm thinking about the approach to create 2 yaml output files:

• velero install --crds-only --dry-run -o yaml > debug_velero.yml
• kubectl apply --dry-run=client -f crds/ -o yaml > debug_origin.yaml

Then we compare the debug_velero.yml with debug_origin.yaml to see if they are synced.
Not sure it the right approach, could you please advise and correct if I was wrong? Thanks

[jenting]

@tungbq
Sounds good.

Be aware that the YAML output ordering and some extra labels or annotations added/removed from the Velero CLI output.
Essentially, the Velero helm chart labels and annotations should be the same as Velero CLI output. but I am not 100% sure if fully the same or not right now.

[tungbq]

@tungbq Sounds good.

Be aware that the YAML output ordering and some extra labels or annotations added/removed from the Velero CLI output. Essentially, the Velero helm chart labels and annotations should be the same as Velero CLI output. but I am not 100% sure if fully the same or not right now.

As I observed, the diff part mostly is:

    status:
      acceptedNames:
        kind: ''
        plural: ''
      conditions: []
      storedVersions: []

and

kubectl.kubernetes.io/last-applied-configuration: >
    SOME_EXTRA_METADATA...

Trying to find a way to automate this. Not quite simple than I thought.
If you have any advice, please let me know. Thanks!

[jenting]

I'd suggest not using kubectl --dry-run=client to generate the CRDs YAML.
Instead, tried to find a way to aggregate the CRDs YAML into one file with the same ordering as velero install --crds-only --dry-run -o yaml.

[jenting]

Or as an alternative, you could compare the two results.

velero install --crds-only --dry-run -o yaml | kubectl apply --dry-run=client -f - -oyaml > debug_velero.yml
kubectl apply --dry-run=client -f crds/ -o yaml > debug_origin.yaml

[tungbq]

Thanks @jenting. I'm implementing the workflow upon your suggestion!