This project demonstrates a fully containerized IoT data pipeline, consisting of three Python services—[[Publisher]], [[Subscriber]], and [[Watchdog]]—all orchestrated with [[Docker]] Compose. Each component runs in its own container, communicates over MQTT, persists data in MySQL, and (for Watchdog) sends email alerts. Docker ensures consistency, isolation, and easy deployment. The [[DockerFile]] also makes watchdog.py an image.

A simple Python script that:

• Simulates sensor readings (light, AC temperature, heater temperature) using a normal distribution plus occasional spikes.
• Publishes those readings as JSON messages to MQTT topics:
  • devices/light
  • devices/ac
  • devices/heater
• Interactive controls via keyboard:
  • Press g to start publishing
  • Press s to stop
  • Press q to quit

• generate_data(): returns {"timestamp":…, "<metric>": value}
• publish_data(): sends JSON to the broker and logs to console
• on_connect(): confirms MQTT connection

A Python service that:

• Connects to the same MQTT broker over TLS
• Listens on the three topics above
• Parses each incoming JSON message
• Inserts the timestamp and metric into MySQL tables lights, ac, and heater

• on_connect(): subscribes to topics
• on_message(): routes each message to the correct insert_* function
• create_tables(): ensures tables exist on startup

A Python “watchdog” that:

• Reads the last 10 entries from the lights table every 10 s
• Computes the average light intensity
• Fetches the latest single reading
• Checks if it exceeds 130% of the average
• Sends an email alert via SMTP (Gmail) when the threshold is breached

• get_last_10_readings(), get_latest_reading(), calculate_average()
• send_alert_email(): constructs and sends an alert message
• monitor_and_alert(): main loop tying everything together

1. Publisher Container

   • Runs the publisher script
   • Publishes to MQTT topics on the broker

2. MQTT Broker

   • Acts as the “post office” passing messages from Publisher → Subscriber

3. Subscriber Container

   • Subscribes to these topics
   • Writes each reading into MySQL

4. MySQL Container

   • Persists all sensor readings in durable storage

5. Watchdog Container

   • Reads data from MySQL
   • Performs threshold check
   • Sends email alerts

All services communicate over a private Docker network (app-network), so no external ports are exposed (except for monitoring UIs like phpMyAdmin).

• Isolation: Each service runs in its own container with its dependencies locked down.
• Reproducibility: “It works on my machine” is solved—containers behave identically on any host with Docker.
• Orchestration: Docker Compose brings up all components (docker-compose up -d) and tears them down cleanly (docker-compose down).
• Networking: Containers automatically share a bridge network, making service discovery easy—e.g., subscriber reaches MySQL at hostname mysql.
• Persistence: Named volumes (e.g., mysql_data) ensure database data survives container restarts.
• Scalability: You can spin up multiple publishers or watchdog instances without worrying about port conflicts or environment mismatches.

# Start all services in background
docker-compose up -d

# See container status and mapped ports
docker-compose ps

# View logs in real time
docker-compose logs -f

# Stop and remove containers (volumes remain)
docker-compose down