Access service across clusters within native Service

In Karmada, the MultiClusterService can enable users to access service across clusters with the native Service domain name, like foo.svc, with the aim of providing users with a seamless experience when accessing service across multiple clusters, as if they were operating within a single cluster.

This document provides an example of how to enable MultiClusterService for accessing service across clusters with native Service.

Prerequisites

Karmada has been installed

We can install Karmada by referring to Quick Start, or directly run hack/local-up-karmada.sh script which is also used to run our E2E cases.

Member Cluster Network

Ensure that at least two clusters have been added to Karmada, and the container networks between member clusters are connected.

• If you use the hack/local-up-karmada.sh script to deploy Karmada, Karmada will have three member clusters, and the container networks of the member1 and member2 will be connected.
• You can use Submariner or other related open source projects to connected networks between member clusters.

note

In order to prevent routing conflicts, Pod and Service CIDRs of clusters need non-overlapping.

Enable MultiClusterService in karmada-controller-manager

To enable the MultiClusterService feature in the karmada-controller-manager, run the following command:

kubectl --context karmada-host get deploy karmada-controller-manager -n karmada-system -o yaml | sed '/- --v=4/i \        - --feature-gates=MultiClusterService=true' | kubectl --context karmada-host replace -f -

Please note that the MultiClusterService feature is disabled by default and can be enabled using the --feature-gates=MultiClusterService=true flag.

If you prefer a more cautious approach, follow these steps:

1. Run kubectl --context karmada-host edit deploy karmada-controller-manager -n karmada-system
2. Check if --feature-gates=MultiClusterService=true is present in the spec.template.spec.containers[0].command field.
3. If not, add --feature-gates=MultiClusterService=true to enable the feature.

Deploy deployment in member1 cluster

We need to deploy deployment in member1 cluster:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx
  labels:
    app: nginx
spec:
  replicas: 2
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - image: nginx
        name: nginx
        resources:
          requests:
            cpu: 25m
            memory: 64Mi
          limits:
            cpu: 25m
            memory: 64Mi
---
apiVersion: policy.karmada.io/v1alpha1
kind: PropagationPolicy
metadata:
  name: nginx-propagation
spec:
  resourceSelectors:
    - apiVersion: apps/v1
      kind: Deployment
      name: nginx
  placement:
    clusterAffinity:
      clusterNames:
        - member1

After deploying, you can check the created pods:

$ karmadactl get po
NAME                     CLUSTER   READY   STATUS              RESTARTS   AGE
nginx-5c54b4855f-6sq9s   member1   1/1     Running             0          28s
nginx-5c54b4855f-vp948   member1   1/1     Running             0          28s

Deploy curl pod in member2 cluster

Let's deploy a curl pod in member2 cluster:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: curl
  labels:
    app: curl
spec:
  replicas: 1
  selector:
    matchLabels:
      app: curl
  template:
    metadata:
      labels:
        app: curl
    spec:
      containers:
      - image: curlimages/curl:latest
        command: ["sleep", "infinity"]
        name: curl
        resources:
          requests:
            cpu: 25m
            memory: 64Mi
          limits:
            cpu: 25m
            memory: 64Mi
---
apiVersion: policy.karmada.io/v1alpha1
kind: PropagationPolicy
metadata:
  name: curl-propagation
spec:
  resourceSelectors:
    - apiVersion: apps/v1
      kind: Deployment
      name: curl
  placement:
    clusterAffinity:
      clusterNames:
        - member2

After deploying, you can check the created pods:

$ karmadactl get po -C member2
NAME                    CLUSTER   READY   STATUS    RESTARTS   AGE
curl-6894f46595-c75rc   member2   1/1     Running   0          15s

Later, we will run the curl command in this pod.

Deploy MultiClusterService and Service in Karmada

Now, instead of using PropagationPolicy/ClusterPropagationPolicy for the Service, we utilize MultiClusterService for propagation.

To enable MultiClusterService in Karmada, deploy the following yaml:

apiVersion: v1
kind: Service
metadata:
  name: nginx
spec:
  ports:
  - port: 80
    targetPort: 80
  selector:
    app: nginx
---
apiVersion: networking.karmada.io/v1alpha1
kind: MultiClusterService
metadata:
   name: nginx
spec:
  types:
    - CrossCluster
  consumerClusters:
    - name: member2
  providerClusters:
    - name: member1

Access the backend pods from member2 cluster

To access the backend pods in the member1 cluster from the member2 cluster, execute the following command:

$ karmadactl exec -C member2 curl-6894f46595-c75rc -it -- sh
~ $ curl http://nginx.default
Hello, world!
Version: 1.0.0
Hostname: nginx-0

Using MultiClusterService, the pods are situated solely in the member1 cluster. However, they can be accessed from the member2 cluster using the native Service name.