Hello, I am experiencing an issue when passing values to the ngrok-ingress-controller helm chart. Although when deploying it, I see no error messages. However, when I actually deploy the test application given in the ngrok guide, I receive an error message indicating the following: HTTP 403: The API authentication you specified is properly formed, but it is invalid. [ERR_NGROK_203] After communicating with ngrok support, they are assuming it is due to how I am passing the values to the helm chart when installing the ingress controller. However, from my perspective I believe I have configured everything correctly. Here is the appropriate yaml for the configuration: # ngrok ingress controller installation

kind: Deploy
type: helm
name: install-ngrok
varfiles:
  - ngrok-helm-values.yml
spec:
  chart:
    name: ngrok-ingress-controller
    repo: https://ngrok.github.io/kubernetes-ingress-controller
  valueFiles:
    - ngrok-helm-values.yml

# ngrok-helm-values.yml file

credentials:
  apiKey: 'MY-API-KEY'
  authtoken: 'MY-AUTH-TOKEN'

According to the ngrok installation guide, which can be found here: https://ngrok.com/docs/using-ngrok-with/k8s/ the helm chart should be installed as follows:

helm install ngrok-ingress-controller ngrok/kubernetes-ingress-controller \
  --namespace ngrok-ingress-controller \
  --create-namespace \
  --set credentials.apiKey=$NGROK_API_KEY \
  --set credentials.authtoken=$NGROK_AUTHTOKEN

Is there something I am doing incorrectly with how I have configured it inside of garden that could be causing the credentials to not be set correctly? Any help or support is very much appreciated. Thanks

Hi @swift-spring-8620! This looks good at a glance, so I'm not quite sure what's going on here. Is

ngrok-helm-values.yml

located in the same directory as the

install-ngrok

action? Also, could you run the command at the

silly

log level (e.g. via

garden deploy -l5

)? That makes things very verbose, but somewhere in the output you should see the actual call to

helm

, including all the CLI args & opts, so you should see the value file paths being passed to the

helm

CLI. I'm wondering if the path to the value file might be incorrect, and that this is being silently ignored without an error (for some reason). Hopefully that helps narrow things down a bit.