Home Weather Monitoring System

IoT environmental monitoring using BME280/CCS811 sensors on Raspberry Pi 4 with local Prometheus and Grafana Cloud integration.

Quick Start

# Configure environment
cp .env.example .env
# Edit .env with your Grafana Cloud credentials

# Deploy
docker-compose up -d

Hardware

• Raspberry Pi 4 with I2C enabled
• SparkFun Environmental Combo Breakout - CCS811/BME280 (Qwiic)
• Optional: pHAT shutdown button

Architecture

The system uses a two-tier monitoring approach:

• Local Prometheus collects and stores sensor metrics locally
• Grafana Cloud receives metrics via Prometheus remote_write for cloud monitoring

Benefits:

✅ Local data retention • ✅ Network resilience • ✅ Cloud monitoring • ✅ Local access to metrics

Deployment Options

Standard Deployment

Local Prometheus + Grafana Cloud integration:

1. Sign up at https://grafana.com (free tier available)
2. Get credentials: Cloud Portal > Prometheus > Details
3. Add to .env: GRAFANA_CLOUD_PUSH_URL, GRAFANA_CLOUD_USERNAME, GRAFANA_CLOUD_PASSWORD

Balena Cloud

✅ Fleet management • ✅ OTA updates • ✅ Remote monitoring

balena login
balena app create home-monitoring --type raspberrypi4-64
balena push home-monitoring

Set environment variables in Balena dashboard or via CLI.

Metrics

Environmental (BME280):

• bme280_humidity - Humidity %
• bme280_pressure - Atmospheric pressure
• bme280_celsius - Temperature °C
• bme280_dewpoint_celsius - Dew point

Air Quality (CCS811):

• ccs811_co2 - CO2 levels (ppm)
• ccs811_tvoc - Total VOCs

All metrics include location tag for filtering.

Configuration

Required Environment Variables:

# Device location tag
DEVICE_DB_LOCATION=home

# Grafana Cloud
GRAFANA_CLOUD_PUSH_URL=https://prometheus-prod-XX.grafana.net/api/prom/push
GRAFANA_CLOUD_USERNAME=your_username  
GRAFANA_CLOUD_PASSWORD=your_password

Services & Access

• Prometheus UI: http://device-ip:9090 (local metrics, queries, alerts)
• Sensor Metrics: http://device-ip:8000/metrics (raw Prometheus format)
• Grafana Cloud: Remote dashboards and alerting

Commands

All commands are run from the project root directory:

# View all logs
docker-compose logs

# View specific service logs
docker-compose logs sensor
docker-compose logs prometheus

# Restart services
docker-compose restart

# Stop all services
docker-compose down

Data Flow

┌─────────────────┐    ┌──────────────────┐    ┌─────────────────┐    ┌─────────────────┐
│   BME280/CCS811 │───▶│  Python Scripts  │───▶│  Prometheus     │───▶│  Grafana Cloud  │
│     Sensors     │    │  (HTTP metrics)  │    │  (local DB)     │    │  (remote_write) │
└─────────────────┘    └──────────────────┘    └─────────────────┘    └─────────────────┘

Troubleshooting

Sensor issues:

• Enable I2C: sudo raspi-config > Interface > I2C
• Check wiring and power
• View logs: docker-compose logs sensor
• Test metrics: curl http://localhost:8000/metrics

Prometheus issues:

• Check Prometheus UI at http://device-ip:9090
• Verify targets are UP in Status > Targets
• Check remote_write queue in Status > Runtime

Network issues:

• Verify priv_lan network exists
• Check internet connectivity for Grafana Cloud
• Ensure Docker daemon is running

Balena deployment:

• Privileged containers enabled for GPIO/I2C access
• Host features: procfs, sysfs for system monitoring

---

Built with Python 3.12, Docker, and Balena.io platform. Based on influxdb-client-python and SparkFun Qwiic libraries.

For Balena builds it seems like you have to hard code secrets in the docker file