Federation Concepts

Kubernetes federation enables managing multiple clusters as a single logical unit. It addresses challenges like geographic distribution, high availability across regions, and workload isolation.

Why Federation

Single-cluster deployments hit limits in production scenarios:

• Geographic latency: Users in different regions need nearby clusters.
• Blast radius: A control plane failure should not take down everything.
• Compliance: Data residency laws may require regional clusters.
• Scale: Very large workloads may exceed single-cluster capacity.

KubeFed (Kubernetes Federation v2)

KubeFed propagates resources across member clusters from a host cluster:

# Install KubeFed control plane
helm install kubefed kubefed-charts/kubefed \
  --namespace kube-federation-system --create-namespace

# Join a member cluster
kubefedctl join member-cluster-1 \
  --host-cluster-context=host \
  --cluster-context=member-1

# Check federation status
kubectl get kubefedclusters -n kube-federation-system

# List joined clusters
kubectl get kubefedclusters -n kube-federation-system \
  -o custom-columns=NAME:.metadata.name,READY:.status.conditions[0].status

Federated Resources

KubeFed uses federated type wrappers to distribute resources:

apiVersion: types.kubefed.io/v1beta1
kind: FederatedDeployment
metadata:
  name: myapp
  namespace: production
spec:
  template:
    spec:
      replicas: 3
      selector:
        matchLabels:
          app: myapp
  placement:
    clusters:
      - name: us-east
      - name: eu-west
  overrides:
    - clusterName: eu-west
      clusterOverrides:
        - path: "/spec/replicas"
          value: 5

Challenges

Federation introduces complexity: conflict resolution across clusters, network connectivity between regions, consistent secret distribution, and cross-cluster service discovery all require careful planning. Many organizations opt for simpler multi-cluster patterns using GitOps instead of full federation.