Week 1 Detailed Tasks - Epic 1: Infrastructure & GitOps

Epic: Epic 1 - DORA 2023 Foundation
Week: 1 of 4
Milestone: 1.1 - Local Infrastructure & 1.2 - GitOps Foundation
Duration: 5 working days
Goal: Deploy local K8s cluster with ArgoCD managing platform

📅 Day-by-Day Breakdown

Day 1: Local Kubernetes Cluster Setup

Issues: #1, #2, #3
Estimated Time: 6-8 hours
Goal: Functional 4-node local K8s cluster with ingress and storage

Morning (4 hours): Issue #1 - K8s Cluster

Task 1.1: Create Terraform module (90 min)

# Location: infra/local/cluster/
 
# Create directory structure
mkdir -p infra/local/cluster/{terraform,inspec/controls,scripts}
 
# File: infra/local/cluster/terraform/main.tf

Copilot Prompt for main.tf:

Create a Terraform configuration for deploying a local Kubernetes cluster using kind:

Requirements:
1. Create a kind cluster named "fawkes-local"
2. Configure 4 worker nodes
3. Each node: 4 CPU, 8GB memory
4. Enable ingress (extraPortMappings for ports 80, 443)
5. Configure local-path-provisioner as default StorageClass
6. Output the kubeconfig path

Include:
- variables.tf (cluster_name, node_count, resources_per_node)
- outputs.tf (kubeconfig_path, cluster_name, node_count)
- versions.tf (terraform >= 1.0, required providers)

Use kind_cluster resource from tehcyx/kind provider.
Add comments explaining each configuration block.

Task 1.2: Create InSpec validation (45 min)

# Location: infra/local/cluster/inspec/controls/cluster.rb

Copilot Prompt:

Create an InSpec test suite for validating Kubernetes cluster deployment:

Test cases:
1. Exactly 4 nodes exist in the cluster
2. All nodes have status "Ready"
3. All system pods in kube-system namespace are "Running"
4. A StorageClass named "local-path" exists and is marked as default
5. Kubernetes version is >= 1.28.0
6. No nodes have MemoryPressure or DiskPressure conditions

Use kubectl commands via command() InSpec resource.
Include descriptive test descriptions and impact levels.

Task 1.3: Create deployment script (90 min)

# Location: scripts/deploy-local-cluster.sh

Copilot Prompt:

Create a bash deployment script that:

1. Checks prerequisites:
   - Docker is installed and running
   - kind CLI is installed (>= 0.20.0)
   - kubectl is installed (>= 1.28.0)
   - terraform is installed (>= 1.0.0)
   - inspec is installed (>= 5.0.0)

2. Deployment steps:
   - cd to infra/local/cluster/terraform
   - Run terraform init
   - Run terraform plan
   - Run terraform apply -auto-approve
   - Export KUBECONFIG from terraform output
   - Wait for all nodes to be Ready (timeout 5 minutes)
   - Wait for all system pods Running (timeout 5 minutes)

3. Validation:
   - Run InSpec tests
   - Print cluster info (nodes, pods, storageclasses)

4. Error handling:
   - Trap errors and provide clear messages
   - Offer rollback option on failure
   - Log all output to logs/cluster-deployment.log

Use colors for output (green=success, red=error, yellow=warning).
Include functions for each major step.

Task 1.4: Create documentation (45 min)

# Location: docs/runbooks/local-cluster-setup.md

Copilot Prompt:

Create comprehensive documentation for local K8s cluster setup:

Sections:
1. Overview (what, why, architecture diagram in mermaid)
2. Prerequisites (system requirements, software versions)
3. Quick Start (5-minute getting started)
4. Detailed Setup (step-by-step with commands)
5. Verification (how to validate deployment)
6. Troubleshooting (common issues and solutions)
7. Cluster Management (start, stop, destroy, reset)
8. Advanced Configuration (resource tuning, multi-cluster)
9. References (links to kind docs, k8s docs)

Use clear language, code blocks with bash syntax highlighting.
Include expected output for verification commands.
Add troubleshooting for:
- Docker not running
- Insufficient resources
- Port conflicts
- StorageClass issues

Validation Commands:

# Test Terraform
cd infra/local/cluster/terraform
terraform init
terraform validate
 
# Test deployment script
chmod +x scripts/deploy-local-cluster.sh
./scripts/deploy-local-cluster.sh
 
# Verify cluster
kubectl get nodes
# Expected: 4 nodes, all Ready
 
kubectl get pods -A
# Expected: All Running
 
kubectl get storageclass
# Expected: local-path (default)
 
# Run InSpec tests
inspec exec infra/local/cluster/inspec/ -t k8s://
 
# Check resource usage
kubectl top nodes
# Expected: <50% CPU, <50% Memory

Afternoon (2-3 hours): Issues #2 & #3 - Ingress & Storage

Task 2.1: Deploy nginx-ingress (90 min)

# Location: platform/apps/ingress-nginx/
mkdir -p platform/apps/ingress-nginx

Copilot Prompt:

Create ArgoCD Application and Helm values for nginx-ingress-controller:

Files needed:
1. values.yaml - Helm values for ingress-nginx chart
   - Enable metrics for Prometheus
   - Configure for local development (NodePort)
   - Set resource limits (500m CPU, 512Mi memory)
   - Enable default SSL certificate

2. application.yaml - ArgoCD Application manifest
   - Name: ingress-nginx
   - Namespace: ingress-nginx (create if not exists)
   - Source: helm chart ingress-nginx/ingress-nginx
   - Sync policy: automated, self-heal
   - Sync options: CreateNamespace=true

3. test-ingress.yaml - Test ingress resource
   - Host: test.local
   - Path: /
   - Backend: test service (port 80)
   - Include TLS with self-signed cert

4. README.md - Setup and usage instructions

Task 3.1: Configure persistent storage (60 min)

# Location: platform/apps/local-path-storage/

Copilot Prompt:

Create configuration for local-path-provisioner:

Files:
1. application.yaml - ArgoCD Application
   - Deploy rancher/local-path-provisioner
   - Set as default StorageClass
   - Configure storage path: /opt/local-path-provisioner

2. test-pvc.yaml - Test PersistentVolumeClaim
   - Name: test-pvc
   - Size: 1Gi
   - AccessMode: ReadWriteOnce
   - StorageClass: local-path

3. test-pod.yaml - Pod that uses test-pvc
   - Mount PVC at /data
   - Run busybox
   - Create test file in /data

Include cleanup script to delete test resources.

End of Day 1 Validation:

# Cluster healthy
kubectl get nodes
kubectl get pods -A
 
# Ingress working
curl http://localhost:80
# Or: curl http://test.local (if configured in /etc/hosts)
 
# Storage working
kubectl apply -f platform/apps/local-path-storage/test-pvc.yaml
kubectl get pvc
# Should show: Bound
 
# Run acceptance test AT-E1-001
./tests/acceptance/run-test.sh AT-E1-001

Day 2: ArgoCD Deployment & Git Structure

Issues: #5, #6, #7
Estimated Time: 6-8 hours
Goal: ArgoCD managing all platform components via GitOps

Morning (4 hours): Issue #5 - Deploy ArgoCD

Task 5.1: Deploy ArgoCD (90 min)

# Location: platform/apps/argocd/

Copilot Prompt:

Create ArgoCD deployment configuration:

1. install.yaml - ArgoCD installation manifest
   - Use official ArgoCD install manifest (v2.9+)
   - Namespace: argocd
   - High availability disabled (for local)
   - Ingress enabled

2. ingress.yaml - Ingress for ArgoCD UI
   - Host: argocd.local
   - HTTPS with self-signed cert
   - Backend: argocd-server:443

3. initial-admin-secret.yaml (optional)
   - Configure initial admin password
   - Store securely or document retrieval method

4. argocd-cm.yaml - ConfigMap for ArgoCD settings
   - Enable anonymous read access (for local dev)
   - Configure repository credentials
   - Set default app project

5. deploy.sh - Deployment script
   - Apply install.yaml
   - Wait for ArgoCD pods Ready
   - Get initial admin password
   - Configure argocd CLI
   - Login and change password
   - Print ArgoCD URL

Include README with:
- How to access UI
- How to login via CLI
- Initial configuration steps

Task 5.2: Configure ArgoCD CLI (30 min)

# Location: scripts/configure-argocd.sh

Copilot Prompt:

Create script to configure ArgoCD CLI:

1. Install argocd CLI if not present
2. Port-forward to argocd-server (8080:443)
3. Login with credentials
4. Add local cluster
5. Create app project "platform"
6. Set up repository credentials for GitHub
7. Verify configuration

Include error handling and validation steps.

Validation:

# Deploy ArgoCD
kubectl apply -f platform/apps/argocd/install.yaml
kubectl wait --for=condition=Ready pods -n argocd --all --timeout=300s
 
# Access UI
kubectl port-forward -n argocd svc/argocd-server 8080:443 &
# Open browser: https://localhost:8080
 
# CLI login
argocd login localhost:8080 --username admin --password $(kubectl -n argocd get secret argocd-initial-admin-secret -o jsonpath="{.data.password}" | base64 -d)
 
# Verify
argocd cluster list
argocd app list

Afternoon (3-4 hours): Issues #6 & #7 - Git Structure & App-of-Apps

Task 6.1: Create Git repository structure (90 min)

# Location: platform/apps/

Copilot Prompt:

Create comprehensive Git repository structure for platform components:

Directory structure:
platform/
├── apps/
│   ├── argocd/                 # ArgoCD itself
│   ├── ingress-nginx/          # Ingress controller
│   ├── local-path-storage/     # Storage provisioner
│   ├── backstage/              # Developer portal (future)
│   ├── jenkins/                # CI/CD (future)
│   ├── harbor/                 # Container registry (future)
│   ├── sonarqube/              # Code quality (future)
│   ├── prometheus/             # Monitoring (future)
│   ├── grafana/                # Dashboards (future)
│   └── dora-metrics/           # DORA service (future)
├── bootstrap/
│   ├── app-of-apps.yaml        # Root application
│   └── README.md               # Bootstrap instructions
└── README.md                   # Platform apps overview

Each app directory contains:
- application.yaml (ArgoCD Application manifest)
- values.yaml (Helm values if using Helm)
- kustomization.yaml (if using Kustomize)
- README.md (app-specific documentation)

Create placeholder directories for future apps.
Add .gitkeep files to maintain empty directories.

Task 6.2: Create app templates (60 min)

# Location: platform/templates/

Copilot Prompt:

Create reusable templates for ArgoCD Applications:

1. helm-app-template.yaml
   - Template for Helm-based apps
   - Variables: {APP_NAME}, {NAMESPACE}, {CHART_REPO}, {CHART_NAME}
   - Standard sync policy
   - Automated pruning and self-heal

2. kustomize-app-template.yaml
   - Template for Kustomize-based apps
   - Variables: {APP_NAME}, {NAMESPACE}, {PATH}

3. generate-app.sh
   - Script to generate application from template
   - Usage: ./generate-app.sh --name myapp --type helm --chart mychart

Include documentation on template usage.

Task 7.1: Implement app-of-apps pattern (90 min)

# Location: platform/bootstrap/

Copilot Prompt:

Create app-of-apps pattern for managing all platform applications:

1. app-of-apps.yaml - Root ArgoCD Application
   - Name: platform-bootstrap
   - Source: platform/apps/
   - Destination: Multiple namespaces
   - Sync wave: 0 (deploy first)
   - Automated sync with prune
   - Uses ApplicationSet for dynamic app discovery

2. applicationset.yaml - ApplicationSet for auto-discovery
   - Git directory generator
   - Path: platform/apps/*/application.yaml
   - Exclude certain directories (templates, common)
   - Preserve application order with sync waves

3. bootstrap.sh - Bootstrap script
   - Apply app-of-apps.yaml
   - Wait for sync
   - Verify all apps deployed
   - Print status

4. README.md - App-of-apps documentation
   - Explain the pattern
   - How to add new apps
   - Sync wave ordering
   - Troubleshooting

Include sync waves:
- Wave 0: Infrastructure (ArgoCD, ingress, storage)
- Wave 1: Platform services (Backstage, Jenkins)
- Wave 2: Observability (Prometheus, Grafana)
- Wave 3: Applications

End of Day 2 Validation:

# Apply app-of-apps
kubectl apply -f platform/bootstrap/app-of-apps.yaml
 
# Wait for sync
argocd app wait platform-bootstrap --timeout 300
 
# Verify all apps
argocd app list
 
# Check sync status
argocd app get platform-bootstrap
 
# Verify applications in cluster
kubectl get applications -n argocd
 
# Run acceptance test AT-E1-002
./tests/acceptance/run-test.sh AT-E1-002

Day 3: Documentation & Validation

Issue: #8, plus documentation
Estimated Time: 6-8 hours
Goal: Complete documentation and validate Week 1 deliverables

Morning (4 hours): Documentation

Task: Create comprehensive Week 1 documentation

Architecture Diagrams (90 min)

# Location: docs/architecture/diagrams/

Create:

infrastructure-overview.mermaid (C4 Context)
kubernetes-cluster.mermaid (Nodes, pods, services)
gitops-workflow.mermaid (Git → ArgoCD → K8s)
network-topology.mermaid (Ingress, services, pods)

Copilot Prompt:

Create Mermaid diagrams for Fawkes architecture:

1. infrastructure-overview.mermaid
   - C4 Context diagram
   - Show: Developer, Git Repo, K8s Cluster, Platform Components
   - Relationships and data flows

2. kubernetes-cluster.mermaid
   - Show 4-node cluster
   - Control plane components
   - Worker node components
   - Storage and networking

3. gitops-workflow.mermaid
   - Sequence diagram
   - Developer commits → Git push → ArgoCD sync → K8s apply
   - Include rollback flow

4. network-topology.mermaid
   - Show ingress controller
   - Service mesh (if applicable)
   - Pod networking
   - External access points

Runbook Updates (90 min)

Update these runbooks:

docs/runbooks/local-cluster-setup.md
docs/runbooks/argocd-management.md (new)
docs/runbooks/gitops-workflow.md (new)
docs/runbooks/troubleshooting-week1.md (new)

README Updates (60 min)

Update:

platform/apps/README.md
infra/README.md
Root README.md (add Week 1 status)

Afternoon (3-4 hours): Validation & Testing

Task 8.1: Run all acceptance tests (90 min)

# Run AT-E1-001
./tests/acceptance/run-test.sh AT-E1-001
 
# Run AT-E1-002
./tests/acceptance/run-test.sh AT-E1-002
 
# Generate test report
./tests/acceptance/generate-report.sh --week 1

Task 8.2: Resource optimization (60 min)

# Check resource usage
kubectl top nodes
kubectl top pods -A
 
# Optimize if needed:
# - Reduce replica counts
# - Adjust resource requests/limits
# - Enable pod disruption budgets
 
# Document optimizations in:
# docs/optimization/week1-resource-tuning.md

Task 8.3: Create demo video (90 min)

Record screen capture demonstrating:

Cluster deployment (2 min)
ArgoCD UI tour (3 min)
Deploying an app via GitOps (5 min)
Viewing synced resources (2 min)
Rollback demonstration (3 min)

Upload to: docs/videos/week1-demo.mp4

End of Day 3 Validation:

# All tests passing
./tests/acceptance/run-all-week1.sh
 
# Documentation complete
find docs/ -name "*.md" | wc -l  # Should be increased
 
# Resource usage acceptable
kubectl top nodes  # All <70%
 
# Demo video created
test -f docs/videos/week1-demo.mp4

Days 4-5: Buffer & Week 2 Prep

Use for:

Catching up on any delayed tasks
Fixing bugs discovered during testing
Additional documentation
Preparing for Week 2 (Backstage, Jenkins)
Team review (if applicable)

✅ Week 1 Definition of Done

📊 Progress Tracking

Update daily in PROJECT_STATUS.md:

### Week 1 Progress
 
**Day 1**: ✅ Cluster deployment complete
- Issue #1: ✅ Complete
- Issue #2: ✅ Complete
- Issue #3: ✅ Complete
 
**Day 2**: ✅ ArgoCD and GitOps complete
- Issue #5: ✅ Complete
- Issue #6: ✅ Complete
- Issue #7: ✅ Complete
 
**Day 3**: ✅ Documentation and validation
- Issue #8: ✅ Complete
- All Week 1 docs: ✅ Complete
 
**Days 4-5**: Buffer used for [describe what was done]
 
**Week 1 Status**: ✅ COMPLETE

🚀 Handoff to Week 2

Once Week 1 is complete, you’re ready for:

Week 2 Focus: Developer Portal & CI/CD

Issue #9: Deploy Backstage
Issue #14: Deploy Jenkins
Issue #17: Deploy Harbor

Prerequisites Met:

✅ K8s cluster operational
✅ GitOps workflow established
✅ Can deploy new apps via ArgoCD

Start Week 2 with:

# Begin Backstage deployment
argocd app create backstage \
  --repo https://github.com/paruff/fawkes \
  --path platform/apps/backstage \
  --dest-server https://kubernetes.default.svc \
  --dest-namespace backstage \
  --sync-policy automated

🆘 Getting Help

If stuck on any task:

Check docs: Search docs/ for related content
Review ADRs: See why decisions were made
Test commands: Validation commands in each task
New chat: Start fresh chat with specific question
GitHub search: Look for similar issues/PRs

Chat template:

Working on Fawkes Epic 1, Week 1, Day X.

Current task: Issue #Y - [Title]
Problem: [Specific issue]
What I've tried: [Commands/approaches]
Error message: [If applicable]

Context: See Week 1 Detailed Tasks document.

END OF WEEK 1 DETAILED TASKS

🪴 PAR Garden

Explorer

week 1 issues

Week 1 Detailed Tasks - Epic 1: Infrastructure & GitOps

📅 Day-by-Day Breakdown

Day 1: Local Kubernetes Cluster Setup

Morning (4 hours): Issue #1 - K8s Cluster

Afternoon (2-3 hours): Issues #2 & #3 - Ingress & Storage

Day 2: ArgoCD Deployment & Git Structure

Morning (4 hours): Issue #5 - Deploy ArgoCD

Afternoon (3-4 hours): Issues #6 & #7 - Git Structure & App-of-Apps

Day 3: Documentation & Validation

Morning (4 hours): Documentation

Afternoon (3-4 hours): Validation & Testing

Days 4-5: Buffer & Week 2 Prep

✅ Week 1 Definition of Done

📊 Progress Tracking

🚀 Handoff to Week 2

🆘 Getting Help

Graph View

Table of Contents

Backlinks