I updated my values.yaml and ran helm upgrade, but I’m seeing the following when watching my cluster’s events with kubectl get events --all-namespaces --watch (showing the webhook-service below, but seeing the same error for all images):
Failed to pull image "speckle/speckle-webhook-service:2.23.5": rpc error: code = DeadlineExceeded desc = failed to pull and unpack image "docker.io/speckle/speckle-webhook-service:2.23.5": failed to resolve reference "docker.io/speckle/speckle-webhook-service:2.23.5": failed to do request: Head "https://registry-1.docker.io/v2/speckle/speckle-webhook-service/manifests/2.23.5": dial tcp 54.198.86.24:443: i/o timeout
When I try to navigate to registry-1.docker.io… I see the following on the page:
{"errors":[{"code":"UNAUTHORIZED","message":"authentication required","detail":[{"Type":"repository","Class":"","Name":"speckle/speckle-webhook-service","Action":"pull"}]}]}.
On the very first helm upgrade, I had docker_image_tag set to a very old value: 2.3.6-alpha.274806, (I think I grabbed the values.yaml from a very old branch on Github). Is it possible that now there’s some sort of caching issue? I’m only now watching the events, so not sure what occurred the first time I deployed the helm chart. Also, I have tried just using docker_image_tag of '2' in ```values.yaml````.
I’m still working with a cluster on AWS EKS, and am also wondering if maybe there’s an IAM policy issue where the nodes don’t have permission to access an external url, but based on my sleuthing online so far, I think that’s unlikely.