Deployment Methods

Greenfield Cluster supports multiple deployment methods to suit different workflows and use cases. Choose the method that best fits your team's needs and infrastructure management approach.

Overview

Method	Best For	Complexity	Customization	GitOps Support
Kustomize	GitOps workflows, precise control	Medium	Excellent	✅ Native
Helm	Quick deployment, parameter-driven	Low	Good	✅ Via ArgoCD/Flux
Direct kubectl	Learning, testing	Low	Limited	❌ Manual

Deployment Method Comparison

Kustomize (Recommended)

Best for: - Teams practicing GitOps - Need for environment-specific configurations - Precise control over Kubernetes manifests - Complex multi-environment deployments

Advantages: - ✅ Native Kubernetes support (built into kubectl) - ✅ Declarative configuration management - ✅ Environment-specific overlays - ✅ Patch-based customization - ✅ No templating language to learn - ✅ Perfect for GitOps (ArgoCD, Flux)

Disadvantages: - ⚠️ Steeper learning curve than Helm - ⚠️ More verbose than templating - ⚠️ Requires understanding of overlay structure

Learn more: Kustomize Deployment Guide

Helm

Best for: - Quick deployments - Parameter-driven configuration - Existing Helm infrastructure - Package distribution

Advantages: - ✅ Simple value-based configuration - ✅ One-command deployment - ✅ Built-in rollback support - ✅ Release management - ✅ Easy to share and distribute

Disadvantages: - ⚠️ Go templating can be complex - ⚠️ Less precise than Kustomize - ⚠️ Chart maintenance overhead

Learn more: Helm Deployment Guide

Direct kubectl

Best for: - Learning and experimentation - Quick testing - Simple deployments

Advantages: - ✅ Simplest to understand - ✅ Direct control - ✅ No additional tools

Disadvantages: - ⚠️ No configuration management - ⚠️ Manual updates required - ⚠️ Not suitable for production - ⚠️ No environment-specific configuration

Quick Start Examples

Using Kustomize

Deploy base configuration:

kubectl apply -k kustomize/base/

Deploy environment-specific overlay:

# Development
kubectl apply -k kustomize/overlays/dev/

# Staging
kubectl apply -k kustomize/overlays/staging/

# Production
kubectl apply -k kustomize/overlays/prod/

Using Helm

Basic deployment:

helm install greenfield helm/greenfield-cluster \
  --namespace greenfield \
  --create-namespace

With custom values:

helm install greenfield helm/greenfield-cluster \
  --namespace greenfield \
  --create-namespace \
  --values my-values.yaml

Using Direct kubectl

Apply individual components:

kubectl apply -f kustomize/base/namespace/namespace.yaml
kubectl apply -f kustomize/base/redis/
kubectl apply -f kustomize/base/postgres/

Environment Management

Development Environment

Characteristics: - Minimal resource requirements - Single replicas for most components - Relaxed resource limits - Suitable for local testing

Kustomize:

kubectl apply -k kustomize/overlays/dev/

Helm:

helm install greenfield helm/greenfield-cluster \
  --values helm/greenfield-cluster/values-dev.yaml

Staging Environment

Characteristics: - Production-like setup - Medium resource allocation - Some redundancy - Pre-production testing

Kustomize:

kubectl apply -k kustomize/overlays/staging/

Helm:

helm install greenfield helm/greenfield-cluster \
  --values helm/greenfield-cluster/values-staging.yaml

Production Environment

Characteristics: - High availability - Multiple replicas - Resource guarantees - Full observability

Kustomize:

kubectl apply -k kustomize/overlays/prod/

Helm:

helm install greenfield helm/greenfield-cluster \
  --values helm/greenfield-cluster/values-prod.yaml

GitOps Integration

ArgoCD

Kustomize Application:

apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
  name: greenfield-cluster
  namespace: argocd
spec:
  project: default
  source:
    repoURL: https://github.com/yourusername/green_field_cluster
    targetRevision: main
    path: kustomize/overlays/prod
  destination:
    server: https://kubernetes.default.svc
    namespace: greenfield
  syncPolicy:
    automated:
      prune: true
      selfHeal: true

Helm Application:

apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
  name: greenfield-cluster
  namespace: argocd
spec:
  project: default
  source:
    repoURL: https://github.com/yourusername/green_field_cluster
    targetRevision: main
    path: helm/greenfield-cluster
    helm:
      valueFiles:
        - values-prod.yaml
  destination:
    server: https://kubernetes.default.svc
    namespace: greenfield
  syncPolicy:
    automated:
      prune: true
      selfHeal: true

Flux CD

Kustomization:

apiVersion: kustomize.toolkit.fluxcd.io/v1beta2
kind: Kustomization
metadata:
  name: greenfield-cluster
  namespace: flux-system
spec:
  interval: 10m
  path: ./kustomize/overlays/prod
  prune: true
  sourceRef:
    kind: GitRepository
    name: green-field-cluster

HelmRelease:

apiVersion: helm.toolkit.fluxcd.io/v2beta1
kind: HelmRelease
metadata:
  name: greenfield-cluster
  namespace: flux-system
spec:
  interval: 10m
  chart:
    spec:
      chart: ./helm/greenfield-cluster
      sourceRef:
        kind: GitRepository
        name: green-field-cluster
  values:
    # Override values here

Component Selection

Removing Unused Components

Kustomize: Edit kustomize/base/kustomization.yaml:

resources:
  - namespace
  - redis
  - postgres
  # - mysql      # Not needed
  # - mongodb    # Not needed
  - kafka
  - istio
  - otel-collector
  - jaeger
  - prometheus
  - grafana

Helm:

# values.yaml
mysql:
  enabled: false
mongodb:
  enabled: false

Adding Custom Components

Kustomize: 1. Create component directory in kustomize/base/ 2. Add component to kustomization.yaml 3. Apply overlay-specific customizations

Helm: 1. Add subchart to helm/greenfield-cluster/charts/ 2. Configure in values.yaml 3. Deploy with updated chart

Upgrading Deployments

Kustomize Upgrades

# Pull latest changes
git pull origin main

# Review changes
kubectl diff -k kustomize/overlays/prod/

# Apply updates
kubectl apply -k kustomize/overlays/prod/

Helm Upgrades

# Update repository
git pull origin main

# Review changes
helm diff upgrade greenfield helm/greenfield-cluster \
  --namespace greenfield

# Upgrade release
helm upgrade greenfield helm/greenfield-cluster \
  --namespace greenfield \
  --values my-values.yaml

Rollback Procedures

Kustomize Rollback

# Revert to previous Git commit
git revert HEAD
git push

# Or manually apply previous version
git checkout <previous-commit>
kubectl apply -k kustomize/overlays/prod/
git checkout main

Helm Rollback

# List release history
helm history greenfield -n greenfield

# Rollback to previous release
helm rollback greenfield -n greenfield

# Rollback to specific revision
helm rollback greenfield 3 -n greenfield

Validation and Testing

Pre-Deployment Validation

Kustomize:

# Generate and review manifests
kubectl kustomize kustomize/overlays/prod/ > output.yaml

# Dry-run deployment
kubectl apply -k kustomize/overlays/prod/ --dry-run=client

# Server-side dry-run
kubectl apply -k kustomize/overlays/prod/ --dry-run=server

Helm:

# Template and review
helm template greenfield helm/greenfield-cluster \
  --values my-values.yaml > output.yaml

# Dry-run
helm install greenfield helm/greenfield-cluster \
  --dry-run --debug

Post-Deployment Validation

# Check pod status
kubectl get pods -n greenfield

# Wait for all pods to be ready
kubectl wait --for=condition=ready pod \
  --all -n greenfield --timeout=600s

# Check service endpoints
kubectl get svc -n greenfield

# View resource usage
kubectl top pods -n greenfield

Cloud Provider Deployment

Deploy to specific cloud providers with platform-specific configurations:

AWS EKS: AWS EKS Deployment Guide
GCP GKE: GCP GKE Deployment Guide
Azure AKS: Azure AKS Deployment Guide

Troubleshooting

Common Issues

Namespace already exists:

# Skip namespace creation or delete existing
kubectl delete namespace greenfield
kubectl apply -k kustomize/overlays/prod/

Storage class not found:

# List available storage classes
kubectl get storageclass

# Update manifests with correct storage class
# Edit PVC definitions in component directories

ImagePullBackOff:

# Check image names and tags
kubectl describe pod <pod-name> -n greenfield

# Ensure image registry is accessible
# Build and push images if using custom builds

Resource constraints:

# Check node resources
kubectl top nodes

# Scale down non-essential components
# Adjust resource requests/limits

Best Practices

Use Version Control: Always commit configuration changes to Git
Environment Isolation: Use separate namespaces or clusters for dev/staging/prod
Secret Management: Use Sealed Secrets or external secret managers
Resource Limits: Always set resource requests and limits
Health Checks: Configure readiness and liveness probes
Monitoring: Deploy observability stack before applications
Backup: Regular backups of persistent data
Documentation: Document custom configurations and changes

Next Steps

Kustomize Deployment Guide - Detailed Kustomize instructions
Helm Deployment Guide - Complete Helm deployment guide
Cloud Provider Guides - Platform-specific deployment
Security Configuration - Secure your deployment