Orchestr8 Architecture
Overview
Orchestr8 (O8) implements a GitOps-based "App of Apps" pattern for managing enterprise Kubernetes applications. This architecture enables:
- Namespace Isolation: Each module runs in its own namespace
- Service Independence: Modules can be deployed, scaled, and updated independently
- Compliance Ready: Built-in support for SOC2, HIPAA, and other compliance frameworks
- Enterprise Scale: Designed to handle hundreds of microservices across multiple clusters
Core Principles
1. GitOps Everything
- All configuration is stored in Git
- Changes are made through Pull Requests
- ArgoCD automatically syncs desired state
2. Module Isolation
- Each module has its own namespace
- Network policies enforce communication boundaries
- RBAC limits access between modules
3. Security First
- Pod Security Standards enforced by default
- Network policies for all modules
- Secrets managed through Sealed Secrets or External Secrets Operator
4. Observability Built-in
- Prometheus metrics for all services
- Distributed tracing with Jaeger
- Centralized logging with Fluentd
Architecture Layers
Platform Layer
┌─────────────────────────────────────────┐
│ ArgoCD (GitOps Engine) │
├─────────────────────────────────────────┤
│ Istio Service Mesh (Traffic Mgmt) │
├─────────────────────────────────────────┤
│ Security & Compliance (Policies) │
├─────────────────────────────────────────┤
│ Observability (Metrics, Logs) │
└─────────────────────────────────────────┘
Application Layer
┌──────────────┐ ┌──────────────┐ ┌──────────────┐
│ VoiceFuse │ │ Langfuse │ │ Your App │
│ Namespace │ │ Namespace │ │ Namespace │
└──────────────┘ └──────────────┘ └──────────────┘
Module Communication
Modules communicate through:
- Service Mesh: Istio manages inter-service communication
- API Gateway: External traffic routes through ingress
- Event Bus: Async communication via NATS/Kafka
- Shared Data: PostgreSQL with schema isolation
Deployment Flow
- Developer commits changes to module repository
- CI/CD builds and pushes container images
- ArgoCD detects changes and syncs
- Kubernetes applies new configuration
- Service mesh handles traffic shifting
Disaster Recovery
- Backup: Velero backs up all resources
- Multi-Region: Supports active-active deployments
- Failover: Automatic failover with health checks
- Data Recovery: Point-in-time recovery for databases