Veritas Kubernetes Manifests

This directory contains Kubernetes manifests for deploying the Veritas system on Google Kubernetes Engine (GKE).

Overview

The Veritas system consists of the following components:

• API: Backend API service for data processing and analysis
• Frontend: Web interface for visualizing and interacting with the data
• Memgraph: Graph database for storing and querying narrative data
• Redis: In-memory cache for performance optimization
• Kafka: Message broker for event-driven architecture

Prerequisites

Before deploying these manifests, ensure you have:

1. A running GKE cluster (deployed via Terraform)
2. kubectl configured to connect to your cluster
3. Container images built and pushed to Google Container Registry (GCR)
4. NGINX Ingress Controller installed (optional, for external access)
5. cert-manager installed (optional, for TLS)

Deployment

1. Update Image References

Before deploying, update the image references in the manifests to point to your GCR repository:

# Replace PROJECT_ID with your actual GCP project ID
sed -i 's/gcr.io\/PROJECT_ID/gcr.io\/your-project-id/g' veritas-deployment.yaml

2. Create Secrets

Create a Kubernetes secret for sensitive information:

kubectl create secret generic veritas-secrets \
  --namespace veritas \
  --from-literal=MEMGRAPH_PASSWORD=your-memgraph-password \
  --from-literal=REDIS_AUTH=your-redis-auth-string \
  --from-literal=JWT_SECRET=your-jwt-secret

3. Deploy the Manifests

Deploy the manifests using kubectl:

kubectl apply -f veritas-deployment.yaml

Or use our deployment script:

../terraform/scripts/deploy-k8s.sh --project-id your-project-id --apply-manifests

4. Verify the Deployment

Check that all pods are running:

kubectl get pods -n veritas

Check the services:

kubectl get services -n veritas

Check the ingress:

kubectl get ingress -n veritas

Configuration

ConfigMaps

The veritas-config ConfigMap contains configuration for the application. You can modify it to suit your needs:

kubectl edit configmap veritas-config -n veritas

Scaling

The deployments include HorizontalPodAutoscalers (HPAs) that automatically scale the pods based on CPU and memory usage. You can modify the scaling parameters:

kubectl edit hpa veritas-api-hpa -n veritas
kubectl edit hpa veritas-frontend-hpa -n veritas

Resource Limits

The resource requests and limits are set to reasonable defaults. You can adjust them based on your workload:

kubectl edit deployment veritas-api -n veritas
kubectl edit deployment veritas-frontend -n veritas
kubectl edit statefulset memgraph -n veritas

Customization

Ingress

The Ingress resource is configured to use NGINX Ingress Controller and Let's Encrypt for TLS. Update the host and annotations as needed:

kubectl edit ingress veritas-ingress -n veritas

Storage

The Memgraph StatefulSet uses a PersistentVolumeClaim for data storage. You can adjust the storage size and class:

# Note: Changing storage for an existing StatefulSet requires additional steps
kubectl edit statefulset memgraph -n veritas

Troubleshooting

Checking Logs

View logs for a specific pod:

kubectl logs -f <pod-name> -n veritas

Debugging

Execute a shell in a pod for debugging:

kubectl exec -it <pod-name> -n veritas -- /bin/sh

Common Issues

1. Pods not starting: Check events with kubectl get events -n veritas
2. Service not accessible: Check service endpoints with kubectl get endpoints -n veritas
3. Ingress not working: Check ingress controller logs and configuration

Maintenance

Updating Images

To update the application to a new version:

kubectl set image deployment/veritas-api api=gcr.io/your-project-id/veritas-api:new-tag -n veritas
kubectl set image deployment/veritas-frontend frontend=gcr.io/your-project-id/veritas-frontend:new-tag -n veritas

Backup and Restore

For Memgraph data, you can create a backup:

# Create a backup job
kubectl apply -f memgraph-backup-job.yaml

Next Steps

After deploying the Veritas system, you can:

1. Set up monitoring with Prometheus and Grafana
2. Configure logging with Stackdriver or EFK stack
3. Implement CI/CD pipelines for automated deployment
4. Set up regular backups for persistent data