5-Layer Model
FaultRay organizes infrastructure components into five logical layers, each representing a distinct tier of the technology stack.
Layer Overview
┌─────────────────────────────────┐
│ Layer 5: Edge / CDN │
├─────────────────────────────────┤
│ Layer 4: DNS / Routing │
├─────────────────────────────────┤
│ Layer 3: Network / Load Bal. │
├─────────────────────────────────┤
│ Layer 2: Compute / Application │
├─────────────────────────────────┤
│ Layer 1: Storage / Database │
└─────────────────────────────────┘
Layer 1: Storage / Database
The foundation layer handles persistent data storage.
- Databases (RDS, Cloud SQL, Cosmos DB)
- Object storage (S3, GCS, Blob Storage)
- File systems (EFS, Filestore)
- Cache layers (ElastiCache, Memorystore)
Key resilience factors: Replication, backup frequency, cross-region failover.
Layer 2: Compute / Application
The application processing layer runs your business logic.
- Virtual machines (EC2, Compute Engine, Azure VMs)
- Containers (ECS, GKE, AKS)
- Serverless (Lambda, Cloud Functions, Azure Functions)
- Application servers
Key resilience factors: Auto-scaling, health checks, rolling deployments.
Layer 3: Network / Load Balancing
The networking layer routes traffic between components.
- Load balancers (ALB/NLB, Cloud Load Balancing, Azure LB)
- VPCs and subnets
- Security groups and firewalls
- Service mesh (Istio, Linkerd)
Key resilience factors: Multi-AZ distribution, health check intervals, connection draining.
Layer 4: DNS / Routing
The routing layer manages name resolution and traffic policies.
- DNS services (Route 53, Cloud DNS, Azure DNS)
- Traffic management (Global Accelerator, Traffic Manager)
- API gateways
Key resilience factors: TTL settings, failover routing policies, health-checked endpoints.
Layer 5: Edge / CDN
The edge layer handles content delivery and DDoS protection.
- CDN (CloudFront, Cloud CDN, Azure CDN)
- WAF / DDoS protection (Shield, Cloud Armor)
- Edge functions (CloudFront Functions, Workers)
Key resilience factors: PoP distribution, origin failover, cache hit ratios.
Cross-Layer Analysis
FaultRay evaluates failure impact across layers. A Layer 1 (database) failure might cascade through Layer 2 (application) and ultimately affect Layer 5 (CDN cache invalidation). The simulation engine traces these cross-layer dependencies to calculate true impact.
faultray simulate -m model.json --cascade-depth 5