DEV Community

Cover image for How to use MQTT in Flask
EMQ Technologies for EMQ Technologies

Posted on • Edited on

How to use MQTT in Flask

Flask is a lightweight Web application framework written with Python, which is called "micro-framework" because it uses a simple core for extension of other features, such as: ORM, form validation tools, file upload, various open authentication techniques, etc.

MQTT is a lightweight Internet of Things (IoT) message transmission protocol based on publish/subscribe mode. It can provide a real-time and reliable message service for networked devices with very less code and smaller bandwidth. It is widely used in IoT, mobile Internet, intelligent hardware, IoV, power and energy industries, etc.

This article mainly introduces how to use MQTT in the Flask project, and implement the connection, subscription, messaging, unsubscribing and other functions between the MQTT client and MQTT broker

We will use the Flask-MQTT client library, which is a Flask extension and can be regarded as a decorator of paho-mqtt to simplify the MQTT integration in Flask applications.

Project Initialization

This project is developed and tested with Python 3.8, and users may use the following commands to verify the version of Python.

$ python3 --version
Python 3.8.2
Enter fullscreen mode Exit fullscreen mode

Use Pip to install the Flask-MQTT library.

pip3 install flask-mqtt
Enter fullscreen mode Exit fullscreen mode

Use Flask-MQTT

We will adopt the Free public MQTT broker provided by EMQ, which is created on the basis of MQTT cloud service - EMQX Cloud. The following is the server access information:

  • Broker: broker.emqx.io
  • TCP Port: 1883
  • Websocket Port: 8083

Import Flask-MQTT

Import the Flask library and Flask-MQTT extension, and create the Flask application.

from flask import Flask, request, jsonify
from flask_mqtt import Mqtt

app = Flask(__name__)
Enter fullscreen mode Exit fullscreen mode

Configure Flask-MQTT extension

app.config['MQTT_BROKER_URL'] = 'broker.emqx.io'
app.config['MQTT_BROKER_PORT'] = 1883
app.config['MQTT_USERNAME'] = ''  # Set this item when you need to verify username and password
app.config['MQTT_PASSWORD'] = ''  # Set this item when you need to verify username and password
app.config['MQTT_KEEPALIVE'] = 5  # Set KeepAlive time in seconds
app.config['MQTT_TLS_ENABLED'] = False  # If your server supports TLS, set it True
topic = '/flask/mqtt'

mqtt_client = Mqtt(app)
Enter fullscreen mode Exit fullscreen mode

For complete configuration items, please refer to Flask-MQTT configuration document.

Write connect callback function

We can handle successful or failed MQTT connections in this callback function, and this example will subscribe to the /flask/mqtt topic after a successful connection.

@mqtt_client.on_connect()
def handle_connect(client, userdata, flags, rc):
   if rc == 0:
       print('Connected successfully')
       mqtt_client.subscribe(topic) # subscribe topic
   else:
       print('Bad connection. Code:', rc)
Enter fullscreen mode Exit fullscreen mode

Write message callback function

This function will print the messages received by the /flask/mqtt topic.

@mqtt_client.on_message()
def handle_mqtt_message(client, userdata, message):
   data = dict(
       topic=message.topic,
       payload=message.payload.decode()
  )
   print('Received message on topic: {topic} with payload: {payload}'.format(**data))
Enter fullscreen mode Exit fullscreen mode

Create message publish API

We create a simple POST API to publish the MQTT messages.

In practical case, the API may need some more complicated business logic processing.

@app.route('/publish', methods=['POST'])
def publish_message():
   request_data = request.get_json()
   publish_result = mqtt_client.publish(request_data['topic'], request_data['msg'])
   return jsonify({'code': publish_result[0]})
Enter fullscreen mode Exit fullscreen mode

Run Flask application

When the Flask application is started, the MQTT client will connect to the server and subscribe to the topic /flask/mqtt.

if __name__ == '__main__':
   app.run(host='127.0.0.1', port=5000)
Enter fullscreen mode Exit fullscreen mode

Test

Now, we use the MQTT client - MQTT X to connect, subscribe, and publish tests.

Receive message

  1. Create a connection in MQTT X and connect to the MQTT server.

MQTT X new connection

  1. Publish Hello from MQTT X to the /flask/mqtt topic in MQTT X.

MQTT X publish MQTT message

  1. We will see the message sent by MQTT X in the Flask running window.

Flask receive MQTT message

Publish message

  1. Subscribe to the /flask/mqtt topic in MQTT X.

MQTT X subscribe

  1. Use Postman to call the /publish API: Send the message Hello from Flask to the /flask/mqtt topic.

Postman test

  1. We can see the message sent from Flask in MQTT X.

Flask publish MQTT message

Complete code

from flask import Flask, request, jsonify
from flask_mqtt import Mqtt

app = Flask(__name__)

app.config['MQTT_BROKER_URL'] = 'broker.emqx.io'
app.config['MQTT_BROKER_PORT'] = 1883
app.config['MQTT_USERNAME'] = ''  # Set this item when you need to verify username and password
app.config['MQTT_PASSWORD'] = ''  # Set this item when you need to verify username and password
app.config['MQTT_KEEPALIVE'] = 5  # Set KeepAlive time in seconds
app.config['MQTT_TLS_ENABLED'] = False  # If your broker supports TLS, set it True
topic = '/flask/mqtt'

mqtt_client = Mqtt(app)


@mqtt_client.on_connect()
def handle_connect(client, userdata, flags, rc):
   if rc == 0:
       print('Connected successfully')
       mqtt_client.subscribe(topic) # subscribe topic
   else:
       print('Bad connection. Code:', rc)


@mqtt_client.on_message()
def handle_mqtt_message(client, userdata, message):
   data = dict(
       topic=message.topic,
       payload=message.payload.decode()
  )
   print('Received message on topic: {topic} with payload: {payload}'.format(**data))


@app.route('/publish', methods=['POST'])
def publish_message():
   request_data = request.get_json()
   publish_result = mqtt_client.publish(request_data['topic'], request_data['msg'])
   return jsonify({'code': publish_result[0]})

if __name__ == '__main__':
   app.run(host='127.0.0.1', port=5000)
Enter fullscreen mode Exit fullscreen mode

Limitations

Flask-MQTT is currently not suitable for the use with multiple worker instances. So if you use a WSGI server like gevent or gunicorn make sure you only have one worker instance.

Summary

So far, we have completed a simple MQTT client using Flask-MQTT and can subscribe and publish messages in the Flask application.

Top comments (0)