When going through the setup of a K8 ingress controller, which is documented here
I am unable to get past the step of "Create an Ingress Controller"
During the Helm command step and placing the command in debug mode, I see there is a timeout on one of the steps:
failed pre-install: timed out waiting for the condition
After reviewing the K8 POD logs, I found that the K8 system is unable to connect to the registry due to an auth error. The following output has been modified for security reasons but shows the error
Failed to pull image "myregistry.azurecr.io/jettech/kube-webhook-certgen:v1.5.1@sha256:...90bd8068": [rpc error: code = NotFound desc = failed to pull and unpack image "....azurecr.io/jettech/kube-webhook-certgen@sha256:....9b9e90bd8068": failed to resolve reference "myregistry.azurecr.io/jettech/kube-webhook-certgen@sha256:...190b1dcbcb9b9e90bd8068": ....azurecr.io/jettech/kube-webhook-certgen@sha256:...9b9e90bd8068: not found, rpc error: code = Unknown desc = failed to pull and unpack image "myregistry.azurecr.io/jettech/kube-webhook-certgen@sha256:...dcbcb9b9e90bd8068": failed to resolve reference "myregistry.azurecr.io/jettech/kube-webhook-certgen@sha256:...b9b9e90bd8068": failed to authorize: failed to fetch anonymous token: unexpected status: 401 Unauthorized]
I have verified that the images are located in the container registry based on the "az acr import" command, and that if I do a standard K8 deployment using "kubectl" , the k8 is able to connect to the acr. I have also verified the connection between the cluster and the registry using the following command it works as expected:
az aks check-acr -n <cluster> -g <rg> --acr <acr>
This failure only occurs when using the helm command.
EDIT
After researching this more, I found the following article
https://stackoverflow.com/questions/68949434/installing-nginx-ingress-controller-into-aks-cluster-cant-pull-image-from-azu
It appears there is a problem with the digest. I added/replaced the following in the helm command:
--set controller.image.digest="sha256:e9fb216ace49dfa4a5983b183067e97496e7a8b307d2093f4278cd550c303899" \
--set controller.admissionWebhooks.patch.image.digest="sha256:950833e19ade18cd389d647efb88992a7cc077abedef343fa59e012d376d79b7" \
However, when running the modified helm command, the PODs are in an error state with the following error
unknown flag: --controller-class
I tried setting the env variable CONTROLLER_TAG=v1.0.0, as documented in the article but this does not help
Another solution is to set the version number: 3.36.0 in the command. This is successful but requires a downgraded version
