If you have two separate kubernetes clusters e.g. local instance of Minikube and some other remote kubernetes cluster and you want them to be able to interact with one another, you can use for that purpose so called Services without selectors.
You can treat your remote kubernetes cluster as any other external resource (e.g. database server running on a separate machine/vm) and you can expose it internally to your cluster. Typically we expose something which is running on our k8s cluster externally but in this case we want to expose an external resource internally to our cluster as if it was one more internal resource. From Pods perspective it actually looks like an internal resource, which is available under its local cluster domain, in exactly the same way as internal resources, exposed by simple ClusterIP service.
Let's take a quick look how to achive it. It's very well explained in this section of the official kubernetes docs:
Services most commonly abstract access to Kubernetes Pods, but they
can also abstract other kinds of backends. For example:
- You want to have an external database cluster in production, but in your test environment you use your own databases.
- You want to point your Service to a Service in a different Namespace
or on another cluster.
- You are migrating a workload to Kubernetes. Whilst evaluating the approach, you run only a proportion of your backends in Kubernetes.
In any of these scenarios you can define a Service without a Pod
selector. For example:
apiVersion: v1
kind: Service
metadata:
name: my-service
spec:
ports:
- protocol: TCP
port: 80
targetPort: 9376
Because this Service has no selector, the corresponding Endpoint
object is not created automatically. You can manually map the Service
to the network address and port where it’s running, by adding an
Endpoint object manually:
apiVersion: v1
kind: Endpoints
metadata:
name: my-service
subsets:
- addresses:
- ip: 192.0.2.42
ports:
- port: 9376
If there is only network connectivity between mentioned dev and staging environments you can make use of Services without selectors. If you need your Minikube cluster to be able to connect to something exposed on 192.0.2.42:9376 by your staging cluster, you just need to use this IP in your Endpoints object definition. Service is defined to listen on por 80 so it will be eventually exposed internally to you Minikube Pods under this standard port.
The biggest advantage of such solution is that such Service makes specific external resource available internally to apps runnig in Pods on your local k8s cluster and they can refer to it using its DNS name i.e. service-name within the same namespace or by FQDN <my-service-name>.<namespace>.svc.cluster.local from any namespace, cluster-wide.
If your external resource is available under a domain name you may also consider using ExternalName, which is a special type of service without selector, and in this case you don't have to configure manually any Endpoints objects. Its definition may look like this:
apiVersion: v1
kind: Service
metadata:
name: my-service
namespace: prod
spec:
type: ExternalName
externalName: my.database.example.com
