9.2. Load-balancing with Global Services

This lab will guide you to perform load-balancing and service discovery across multiple Kubernetes clusters.

Task 9.2.1: Load-balancing with Global Services

Establishing load-balancing between clusters is achieved by defining a Kubernetes service with an identical name and Namespace in each cluster and adding the annotation io.cilium/global-service: "true" to declare it global. Cilium will automatically perform load-balancing to pods in both clusters.

We are going to deploy a global service and a sample application on both of our connected clusters.

First the Kubernetes service. Create a file svc.yaml with the following content:

---
apiVersion: v1
kind: Service
metadata:
  name: rebel-base
  annotations:
    io.cilium/global-service: "true"
spec:
  type: ClusterIP
  ports:
  - port: 80
  selector:
    name: rebel-base

Apply this with:

kubectl --context cluster1 apply -f svc.yaml
kubectl --context cluster2 apply -f svc.yaml

Then deploy our sample application on both clusters.

cluster1.yaml:

---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: rebel-base
spec:
  selector:
    matchLabels:
      name: rebel-base
  replicas: 2
  template:
    metadata:
      labels:
        name: rebel-base
    spec:
      containers:
      - name: rebel-base
        image: docker.io/nginx:1.15.8
        volumeMounts:
          - name: html
            mountPath: /usr/share/nginx/html/
        livenessProbe:
          httpGet:
            path: /
            port: 80
          periodSeconds: 1
        readinessProbe:
          httpGet:
            path: /
            port: 80
      volumes:
        - name: html
          configMap:
            name: rebel-base-response
            items:
              - key: message
                path: index.html
---
apiVersion: v1
kind: ConfigMap
metadata:
  name: rebel-base-response
data:
  message: "{\"Galaxy\": \"Alderaan\", \"Cluster\": \"Cluster-1\"}\n"
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: x-wing
spec:
  selector:
    matchLabels:
      name: x-wing
  replicas: 2
  template:
    metadata:
      labels:
        name: x-wing
    spec:
      containers:
      - name: x-wing-container
        image: docker.io/cilium/json-mock:1.2
        livenessProbe:
          exec:
            command:
            - curl
            - -sS
            - -o
            - /dev/null
            - localhost
        readinessProbe:
          exec:
            command:
            - curl
            - -sS
            - -o
            - /dev/null
            - localhost
kubectl --context cluster1 apply -f cluster1.yaml

cluster2.yaml:

---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: rebel-base
spec:
  selector:
    matchLabels:
      name: rebel-base
  replicas: 2
  template:
    metadata:
      labels:
        name: rebel-base
    spec:
      containers:
      - name: rebel-base
        image: docker.io/nginx:1.15.8
        volumeMounts:
          - name: html
            mountPath: /usr/share/nginx/html/
        livenessProbe:
          httpGet:
            path: /
            port: 80
          periodSeconds: 1
        readinessProbe:
          httpGet:
            path: /
            port: 80
      volumes:
        - name: html
          configMap:
            name: rebel-base-response
            items:
              - key: message
                path: index.html
---
apiVersion: v1
kind: ConfigMap
metadata:
  name: rebel-base-response
data:
  message: "{\"Galaxy\": \"Alderaan\", \"Cluster\": \"Cluster-2\"}\n"
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: x-wing
spec:
  selector:
    matchLabels:
      name: x-wing
  replicas: 2
  template:
    metadata:
      labels:
        name: x-wing
    spec:
      containers:
      - name: x-wing-container
        image: docker.io/cilium/json-mock:1.2
        livenessProbe:
          exec:
            command:
            - curl
            - -sS
            - -o
            - /dev/null
            - localhost
        readinessProbe:
          exec:
            command:
            - curl
            - -sS
            - -o
            - /dev/null
            - localhost
kubectl --context cluster2 apply -f cluster2.yaml

Now you can execute from either cluster the following command (there are two x-wing pods, simply select one):

XWINGPOD=$(kubectl --context cluster1 get pod -l name=x-wing -o jsonpath="{.items[0].metadata.name}")
for i in {1..10}; do
  kubectl --context cluster1  exec -it $XWINGPOD -- curl -m 1 rebel-base
done

as a result you get the following output:

{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-1"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-1"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-1"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-1"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-1"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-1"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-1"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}

and as you see, you get results from both clusters. Even if you scale down your rebel-base Deployment on cluster1 with

kubectl --context cluster1 scale deployment rebel-base --replicas=0

and then execute the curl for loop again, you still get answers, this time only from cluster2:

{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}
{"Galaxy": "Alderaan", "Cluster": "Cluster-2"}

Scale your rebel-base Deployment back to one replica:

kubectl --context cluster1 scale deployment rebel-base --replicas=1
Last modified August 6, 2025: Merge pull request #257 from acend/renovate/docker-login-action-digest (92607dc)