Customer is trying to deploy a portion of their application on Kubernetes. The are using Azure / AKS. They have a phased approach to using Kubernetes.
They have an internal ip address that they want to use. They have merged the yaml described in the below Microsoft documentation with our helm chart.
What happens is the deployment fails initially. Then after some time they run kubectl get services, we can see the external IP address gets assigned, but the manager pod is still in PodInitializing state and does not manage to get out of this.
If we do kubectl describe pod <pod name> or kubectl logs <pod name> or kubectl logs --previous <pod name> or kubectl describe deployment, we don't see any information to help us understand why it's in initializing state or we get an error.
This is a showstopper for them as they are trying to meet a release date.
They are using loadBalancer, as this is the default for version 15.0.