Platform Edge

Edge deployment platform for ONA Platform services, optimized for ARM64 devices like Raspberry Pi CM4.

Overview

This repository contains edge-optimized versions of ONA Platform services, designed to run on resource-constrained devices with ARM64 architecture. The primary focus is on deploying the forecasting API service for local inference.

Architecture

• Target Devices: Raspberry Pi CM4, Orange Pi 5, Rock 5B, and similar ARM64 SBCs
• OS Support: Raspberry Pi OS, Ubuntu ARM64, Debian ARM64
• Container Runtime: Docker with ARM64 support
• Services: forecastingApi (primary), with extensibility for other services

Key Features

• ARM64 Optimized: All containers built for ARM64 architecture
• Resource Efficient: Optimized for low-power, low-memory devices
• Edge Native: Designed for local inference without cloud dependencies
• Model Management: Automated model updates and caching
• Monitoring: Built-in health checks and performance monitoring

Quick Start

Prerequisites

• ARM64 device (Raspberry Pi CM4 recommended)
• Docker installed
• 4GB+ RAM (8GB recommended)
• 16GB+ storage
• Internet connection for model updates

Installation

# Clone the repository
git clone <repository-url> platform-edge
cd platform-edge

# Setup sync with main platform repository
./scripts/setup-sync.sh

# Build and start services
docker-compose up -d

# Check status
docker-compose ps

Configuration

1. Copy config/environment.example.sh to config/environment.sh
2. Update configuration variables
3. Restart services: docker-compose restart

Services

forecastingApi

Edge-optimized forecasting service for solar energy prediction.

Features:

• Pre-trained model inference
• Redis caching
• Automated model updates
• RESTful API endpoints

Endpoints:

• GET /health - Health check
• POST /forecast - Generate forecast
• GET /status - Service status

Development

Local Development

# Start development environment
docker-compose -f docker-compose.dev.yml up

# Run tests
make test

# Build for production
make build

Syncing with Main Repository

# Sync service code from main platform repository
./scripts/sync-services.sh

# Check sync status
./scripts/sync-services.sh status

# Validate current state
./scripts/sync-services.sh validate

Adding New Services

1. Add service to SERVICES_TO_SYNC in scripts/sync-services.sh
2. Define sync files in SYNC_FILES array
3. Define preserve files in PRESERVE_FILES array
4. Create edge-specific Dockerfile and configuration
5. Update docker-compose.yml

Deployment

Production Deployment

# Deploy to edge device
./scripts/deploy.sh

# Update models
./scripts/update-models.sh

# Monitor services
./scripts/monitor.sh

Model Management

• Models are automatically downloaded from cloud storage
• Weekly model updates are scheduled
• Local model caching reduces download times
• Model validation ensures integrity

Monitoring

Health Checks

• Service health endpoints
• Resource usage monitoring
• Model freshness tracking
• Error rate monitoring

Logs

# View service logs
docker-compose logs -f forecasting-api

# View all logs
docker-compose logs -f

Performance

Expected Performance (CM4 8GB)

• Forecast Generation: 2-5 seconds
• Memory Usage: 1-2GB
• CPU Usage: <20% peak
• Storage: <1GB for models
• Power: 3-5W active, 1W idle

Troubleshooting

Common Issues

1. Out of Memory: Increase swap or use 8GB CM4
2. Model Download Fails: Check network connectivity
3. Slow Performance: Ensure ARM64 optimized containers
4. Docker Issues: Update Docker to latest version

Support

• Check logs: docker-compose logs
• Verify configuration: ./scripts/check-config.sh
• Test connectivity: ./scripts/test-connectivity.sh

Contributing

1. Fork the repository
2. Create feature branch
3. Make changes
4. Test on ARM64 device
5. Submit pull request

License

[Add your license here]

Changelog

v1.0.0

• Initial release
• forecastingApi edge deployment
• ARM64 optimization
• Docker Compose setup
• Model management system