RESTful API : Application tutorial

This document is a tutorial to use the edge computing API: RESTful API. In the following, you will be shown how to use the API, how to create an application to communicate with the API and how to communicate with the API to get the results of an application.

Introduction

For this tutorial, we will take the following case:

Once the API is launched, a data prediction Application A will start and report its presence on the network to the API and a data generation Application B will send a set of data to the API. Finally, a test user C, will ask the API to launch Application A and return its prediction based on the data sent by Application B.

Installation

First downlaod, install and run the API : https://github.com/llucbono/Edge_v3

docker-compose -f local.yml up

Then install and run the data prediction application A:

docker build -t app_demo_prediction .
docker run -p 5000:5000 -d app_demo_prediction

Go the http://localhost:5000/hi to see if the prediction application A is working fine.

Demo

Then run the data generation application B :

python demoAppDataGenerator.py

The API is now receiving a series of random data at regular intervals.

Then run the test user C :

python demoAppUser.py

The user asks the API for the application's prediction about the data sent just before.

How build an application to communicate with the API ?

Each of your applications should have the following form:

from appInterface import ApplicationInterface
from flask import Flask
from multiprocessing import Process

URL = "URL WHERE THE API IS RUNNING"
LOCAL_IP = "IP OF THE COMPUTER WHERE THIS APP IS RUNNING"
APPNAME="NAME OF THE APP"

interface = ApplicationInterface(URL)
app = Flask(__name__)

def startCommunication(app):
    server = Process(target=app.run(debug= True, port=5000))
    server.start()    

def stopCommunication(server):
    server.terminate()
    server.join()

@app.route('/send-ip')
def send_ip():
    try:
        interface.postIP(LOCAL_IP,'12','appIP',APPNAME)
        return LOCAL_IP
    except:
        return 'DEBUG: Error sending IP'
        
@app.route('/run-app')
def run_app():
    # YOUR CODE HERE
    res = interface.getListOfMessageFromSensorType("deg") # get the list of data from the sensor type "deg"
    data = res['data']
    toDo(data)

def exit_handler():
    stopCommunication(app)
    interface.deleteAppIPbyName(APPNAME)

def toDo(data):
    # YOUR CODE HERE
    pass

def main():
    try:
        interface.postIP(LOCAL_IP,'12','appIP',APPNAME)
        print('[+] IP send to the API', LOCAL_IP)
    except:
        print('DEBUG: Error sending IP')
    startCommunication(app)

if __name__ == '__main__':
    main()

How call your application using the API ?

Below you can find an example of the user side asking to run an application:

from appInterface import ApplicationInterface
import requests

APPNAME="NAME OF THE APP"
API_URL = "URL WHERE THE API IS RUNNING"
interface = ApplicationInterface(API_URL)

appIP = interface.getAppIPbyName(APPNAME)['data']

s = requests.Session()

_appURL = "http://" + appIP + ":5000/run-app"
resp = s.get(url=_appURL)
print('[+] Message from App:',resp.text)

jbaudru / RESTful-app-and-interface