Readme

Monitoring Apartment with Raspberry Pi, Prometheus & Grafana. A Pi Home python workshop based in original pdambrauskas's Pi Home.

Hardware components

• Raspberry Pi 3 Model B
• 16 GB microSD card
• DHT11 Temperature And Humidity Sensor
• Motion Sensor HC-SR501
• Raspberry Pi Camera module.

Python samples

https://gist.github.com/davidalbela/2367febf0d9e7811af54de6ece6040bc

Install

1. Install Adafruit libraries

sudo pip3 install Adafruit_DHT

2. Install Docker & Docker-compose

curl -fsSL https://get.docker.com -o get-docker.sh
sudo sh get-docker.sh

# Give docker rights to Pi user
sudo usermod -aG docker pi

NOTE: Run exit and log in again to make sure usermod works.

3. Install Docker-compose

sudo apt install libffi-dev libssl-dev
sudo pip3 install docker-compose
sudo systemctl enable docker

Create a web service

Go to flask path and follow steps:

1. Create sensor flask_app.py

#!/bin/env python

import Adafruit_DHT
import RPi.GPIO as GPIO
import time

GPIO.setmode(GPIO.BOARD)

sensor = Adafruit_DHT.DHT11

# TODO 
sensor_pin = # Indicar el pin GPIO donde se conecta el sensor

# Add webserver
from flask import Flask
app = Flask(__name__)

# TODO What is `/metrics`
@app.route('/metrics')
def metrics():
    umid, temp = Adafruit_DHT.read_retry(sensor, sensor_pin)
    if umid is not None and temp is not None:
        return '# HELP local_temp local temperature\n# TYPE local_temp gauge\nlocal_temp {}\n# HELP local_humidity local humidity\n# TYPE local_humidity gauge\nlocal_humidity {}\n'.format(int(temp), int(umid)), 200, {'Content-Type': 'text/plain; charset=utf-8'}
    else:
        return 'Could not read from DHT11.', 200, {'Content-Type': 'text/plain; charset=utf-8'}

2. Test flask

# Run the server
export FLASK_APP=/home/pi/flask_app.py
export FLASK_DEBUG=1
flask run --host=0.0.0.0 --port=5000

# Open other terminal and run
curl [pihome IP]:5000/metrics

Why /metrics?

Press Control + c to exit.

3. Create a run script flash.sh file

# Run nano flash.sh
#!/bin/bash

# flask settings
export FLASK_APP=/home/pi/flask_app.py
export FLASK_DEBUG=0

flask run --host=0.0.0.0 --port=5000

# Save and Run chmod +x flask.sh

4. Install flash.sh as a daemon

# Run cd /etc/systemd/system/
# Run sudo nano flask.service
[Unit]
Description = flask python command to do useful stuff

[Service]
ExecStart = /home/pi/flask.sh

[Install]
WantedBy = multi-user.target

# Save file

5. Run and enable flask as a daemon

sudo systemctl start flask
sudo systemctl enable flask.service

NOTE: Check service logs with journal:

journalctl -u flask.service

Execute Monitoring System

1. Clone this repo

git clone https://github.com/davidalbela/pihome.git
cd pihome

2. Setup environment variables in .env file

nano .env
# TODO paste here your Raspberry Pi IP
HOST_IP=
# TODO paste here a secure password for Grafana service
GRAFANA_PASSWORD=

3. Execute Prometheus and Grafana as docker containers

docker-compose up -d

NOTE: To stop containers you can execute docker-compose down.

4. Check your containers

# list containers
docker ps

# read logs
docker logs grafana
docker logs -f prometheus

5. Visit your Raspberry Pi IP in your browser:

http://[Raspberry Pi IP]

Enter the user "admin" and your password configured in .env file (value for GRAFANA_PASSWORD).