Skip to content

K8GB with Google Cloud DNS - Example

This example demonstrates how to configure K8GB with Google Cloud DNS across two GKE clusters in different regions.

📖 For detailed setup instructions, architecture explanation, and troubleshooting, see Google Cloud DNS Provider Documentation

Quick Overview

This example deploys K8GB across two GKE clusters: - Primary cluster: europe-west1 (failover primary) - Secondary cluster: us-central1 (failover secondary) - DNS Zone: gcp-test.k8gb.io (load-balanced zone hosted by CoreDNS) - Parent Zone: k8gb.io (Cloud DNS managed zone for delegation)

Prerequisites

  1. Two GKE clusters in different regions with Workload Identity enabled
  2. A Cloud DNS managed zone (see provider_gcp.md - Prerequisites)
  3. Service account with roles/dns.admin (see provider_gcp.md - Authentication)

Deployment

  1. Deploy K8GB on the first cluster: 

    # Configure kubectl for europe-west1 cluster
gcloud container clusters get-credentials my-cluster-eu \
  --region=europe-west1 --project=my-gke-project

# Install K8GB
helm upgrade --install k8gb k8gb/k8gb \
  --namespace=k8gb --create-namespace \
  --values=k8gb-cluster-gcp-europe-west1.yaml

  2. Deploy K8GB on the second cluster: 

    # Configure kubectl for us-central1 cluster
gcloud container clusters get-credentials my-cluster-us \
  --region=us-central1 --project=my-gke-project

# Install K8GB
helm upgrade --install k8gb k8gb/k8gb \
  --namespace=k8gb --create-namespace \
  --values=k8gb-cluster-gcp-us-central1.yaml

  3. Deploy test application and GSLB resources: 

    # Create test namespace (do this on both clusters)
kubectl apply -f ../../../deploy/crds/test-namespace-ingress.yaml

# Deploy podinfo via Helm (do this on both clusters)
helm repo add podinfo https://stefanprodan.github.io/podinfo
helm upgrade --install frontend --namespace test-gslb \
  -f ../../../deploy/test-apps/podinfo/podinfo-values.yaml \
  --set ui.message="$(kubectl config current-context)" \
  podinfo/podinfo --version 6.9.2

# Apply GSLB configurations (do this on both clusters)
kubectl apply -f test-gslb-failover.yaml
kubectl apply -f test-gslb-roundrobin.yaml

Configuration Files

This example includes: - k8gb-cluster-gcp-europe-west1.yaml - Configuration for EU cluster - k8gb-cluster-gcp-us-central1.yaml - Configuration for US cluster - test-gslb-failover.yaml - Failover strategy example - test-gslb-roundrobin.yaml - Round-robin strategy example

Key Configuration Notes

  • Workload Identity: Configured by default in the example YAML files
  • Static Credentials: Commented out alternative (uncomment env/volumes sections if needed)
  • Cross-Project Setup: Use extraArgs.google-project if DNS zone is in a different project
  • Geo Tags: Must be unique per cluster (europe-west1, us-central1)
  • TXT Ownership: Each cluster needs unique txtPrefix and txtOwnerId

See provider_gcp.md for detailed configuration parameters.

Verification

# Check K8GB status
kubectl get gslb -n test-gslb
kubectl describe gslb test-gslb-failover -n test-gslb

# Verify NS delegation records in Cloud DNS
gcloud dns record-sets list --zone="k8gb-io" --filter="name~'gcp-test.k8gb.io'"

# Test DNS resolution
nslookup failover.gcp-test.k8gb.io 8.8.8.8
nslookup roundrobin.gcp-test.k8gb.io 8.8.8.8

For detailed verification steps and troubleshooting, see: - Verification - Troubleshooting

Testing Failover

# Scale down the primary cluster application
kubectl scale deployment frontend-podinfo --replicas=0 -n test-gslb

# Wait for health check to detect failure (~30-60 seconds)
# Verify DNS now points to secondary cluster
nslookup failover.gcp-test.k8gb.io 8.8.8.8

Additional Resources