Deployment Infrastructure

Deployment Infrastructure

Core Idea: Docker Compose orchestrates platform services (RabbitMQ, PostgreSQL, Redis, Jaeger) with devbox for development and isolated containers for production.

Overview

The platform uses Docker for consistent deployment across development and production:

• Development: docker-compose.yml with devbox container
• Production: Individual service containers with Docker orchestration
• Infrastructure: Shared services (RabbitMQ, PostgreSQL, Redis, Jaeger, Grafana)

Development Deployment

Start Platform

# Start all infrastructure + devbox
docker compose up

# Services started:
# - devbox (development environment)
# - rabbitmq (message bus)
# - postgres (event store)
# - redis (cache)
# - jaeger (tracing)
# - elasticsearch (jaeger backend)
# - grafana (dashboards)

Access Points

• Devbox shell: docker exec -it flow-platform-devbox bash
• RabbitMQ UI: http://localhost:55672 (admin/admin123)
• Jaeger UI: http://localhost:16686
• Grafana: http://localhost:5005 (admin/admin)
• PostgreSQL: localhost:55432 (actor_user/actor_pass123)
• Redis: localhost:56379

Production Deployment

Service Container

FROM node:20-alpine

WORKDIR /app

# Install dependencies
COPY package*.json ./
RUN npm ci --only=production

# Copy application
COPY dist/ ./dist/

# Run service
CMD ["node", "dist/main.js"]

Environment Variables

# Service configuration
NODE_ENV=production
SERVICE_NAME=order-service
SERVICE_VERSION=1.0.0

# Infrastructure
DATABASE_HOST=postgres-production
DATABASE_NAME=eventstore
DATABASE_USER=production_user
DATABASE_PASSWORD=<secure-password>
RABBITMQ_URL=amqp://production-user:<password>@rabbitmq-cluster:5672
REDIS_URL=redis://:<password>@redis-cluster:6379

# Telemetry
JAEGER_ENDPOINT=http://jaeger:4318/v1/traces

Health Checks

healthcheck:
  test: ["CMD", "node", "health-check.js"]
  interval: 30s
  timeout: 10s
  retries: 3
  start_period: 40s

Scaling

Horizontal Scaling

# Scale service instances
docker compose up --scale order-service=3

# Load balancing automatic via message bus
# Each instance consumes from same queue

Resource Limits

services:
  order-service:
    deploy:
      resources:
        limits:
          cpus: '1.0'
          memory: 512M
        reservations:
          cpus: '0.5'
          memory: 256M

Monitoring

Container Health

# Check container status
docker ps

# View logs
docker logs flow-platform-devbox

# Resource usage
docker stats

Infrastructure Health

# RabbitMQ
docker exec rabbitmq rabbitmqctl status

# PostgreSQL
docker exec postgres pg_isready

# Redis
docker exec redis redis-cli ping

Best Practices

1. Use Health Checks

   • All services should expose health endpoint
   • Platform services respond to health check queries

2. Set Resource Limits

   • Prevent resource exhaustion
   • Enable auto-scaling triggers

3. Centralized Logging

   • All logs to stdout/stderr
   • Use logging aggregation (ELK stack)

4. Secrets Management

   • Never commit secrets to repository
   • Use environment variables or secret management

5. Graceful Shutdown

   • Handle SIGTERM signal
   • Complete in-flight requests
   • Close connections cleanly

Troubleshooting

Issue: Service Won’t Start

# Check logs
docker logs <container-name>

# Check environment
docker exec <container-name> env

# Check connectivity
docker exec <container-name> ping postgres

Issue: Database Connection Failed

# Verify PostgreSQL running
docker ps | grep postgres

# Test connection
docker exec postgres pg_isready

# Check credentials
docker exec postgres psql -U actor_user -d eventstore -c "SELECT 1"

Further Reading

• Docker Documentation
• Docker Compose
• Platform Overview