My applications are distributed across several namespaces in a single OpenShift cluster. I want to use two Kong ingress controllers in two separate new name spaces to serve the API traffic from internal and external consumers independently. Both these controllers will use the same set of applications.
What are the recommendations and best practices in designing, installing these Ingress resources/plugins and to integrating with applications running in multiple name spaces.
You can use kubernetes.io/ingress.class with different values and separate out Ingress resources that should be exposed externally and internally.
This also gives you flexibility to expose some applications internally only and some externally only.
If you wish expose each and every application externally and internally, then you can also deploy a single Ingress Controller and then have two services (1 for external, 1 for internal) in Kubernetes that point to the same proxy pods.