Skip to content

deployment

Directory actions

More options

Directory actions

More options

Latest commit

 

History

History

deployment

README.md

Deployment Guides

Comprehensive deployment documentation for various environments and configurations.

Available Guides

Container Deployment

GitHub Actions Runners

Docker Infrastructure

Deployment Options

Container Deployment

  • Docker: Single-container deployment for development
  • Docker Compose: Multi-container orchestration
  • Kubernetes: Cloud-native production deployment

Runner Deployment

  • Standard Runners: x86_64 GitHub Actions runners
  • ARM64 Runners: ARM-based runners for multi-architecture support
  • GPU Runners: CUDA-enabled runners for ML workloads

Production Deployment

  • GraphRAG: Production GraphRAG with knowledge graphs and vector stores
  • MCP Server: Model Context Protocol server (200+ tools)
  • API Servers: RESTful API deployment

Quick Start

Docker Deployment

# Build and run with Docker Compose
docker-compose up -d

# Check status
docker-compose ps

# View logs
docker-compose logs -f

Kubernetes Deployment

# Deploy to Kubernetes
kubectl apply -f deployments/kubernetes/

# Check deployment status
kubectl get pods
kubectl get services

Runner Setup

# Download and configure runner
./config.sh --url https://github.com/your-org/your-repo --token YOUR_TOKEN

# Start runner
./run.sh

Architecture

High Availability Setup

For production HA deployments:

  • Multiple application replicas behind load balancer
  • Redundant IPFS nodes with DHT optimization
  • Replicated vector stores (Qdrant cluster)
  • Database replication (PostgreSQL streaming replication)

Scaling Strategy

  • Horizontal Scaling: Multiple container replicas with auto-scaling
  • Vertical Scaling: Resource allocation per container
  • Database Scaling: Read replicas and connection pooling
  • IPFS Scaling: Gateway caching and DHT optimization

Monitoring & Observability

Essential monitoring setup:

  • Application Metrics: Response times, error rates, throughput
  • IPFS Metrics: Peer count, bandwidth, pin status
  • Resource Metrics: CPU, memory, disk, network
  • Business Metrics: Documents processed, queries executed

See Unified Dashboard for monitoring dashboard setup.

Security

Production security essentials:

  • Network: Firewalls, VPNs, private networks
  • Access Control: API keys, JWT tokens, RBAC
  • Encryption: TLS for transit, encryption at rest
  • Audit Logging: Track all operations

See Security & Governance Guide.

Disaster Recovery

Backup Strategy

  • Configuration files (daily)
  • Vector store indices (incremental)
  • Database dumps (daily with point-in-time recovery)
  • IPFS pinned content lists (daily)

Recovery Procedures

  1. Application Recovery: Redeploy from container registry
  2. Data Recovery: Restore from encrypted backups
  3. IPFS Recovery: Re-pin content from backup lists
  4. Database Recovery: Restore from SQL dumps or PITR

Performance Tuning

Production performance optimization:

  • Enable Redis caching for frequently accessed data
  • Use GPU acceleration for embeddings (2-10x speedup)
  • Optimize IPFS with custom gateway and pin sets
  • Tune vector store parameters (HNSW ef_construction, M)
  • Configure database connection pooling

See Performance Optimization Guide.

Troubleshooting

Common Issues

Deployment fails:

  • Check logs: docker-compose logs or kubectl logs
  • Verify configuration files and secrets
  • Check resource availability (CPU, memory, disk)
  • Validate network connectivity

Poor performance:

  • Review resource allocation and limits
  • Check database query performance
  • Monitor IPFS peer connections
  • Profile application with py-spy or cProfile

Connection errors:

  • Verify network connectivity between services
  • Check firewall rules and security groups
  • Validate DNS resolution
  • Test service endpoints manually

Related Documentation