Scalable & Event Driven simple ticket buying system with Kafka & Python

Apache Kafka is a distributed event store and stream-processing platform. It is an open-source system developed by the Apache Software Foundation written in Java and Scala. The project aims to provide a unified, high-throughput, low-latency platform for handling real-time data feeds.

We are gonna build a simple ticket buying system just to understand basic consumer and producer functionality in Kafka

First of all we need to setup kafka, zookeeper and python kafka package.

For kafka and zookeeper I'm using docker for installation. Here is the docker-compose.yml file

version: '2'
services:
  zookeeper:
    image: confluentinc/cp-zookeeper:latest
    environment:
      ZOOKEEPER_CLIENT_PORT: 2181
      ZOOKEEPER_TICK_TIME: 2000
    ports:
      - 22181:2181
  kafka:
    image: confluentinc/cp-kafka:5.3.1
    depends_on:
      - zookeeper
    ports:
      - 29092:29092
    environment:
      KAFKA_BROKER_ID: 1
      KAFKA_ZOOKEEPER_CONNECT: zookeeper:2181
      KAFKA_ADVERTISED_LISTENERS: PLAINTEXT://kafka:9092,PLAINTEXT_HOST://localhost:29092
      KAFKA_LISTENER_SECURITY_PROTOCOL_MAP: PLAINTEXT:PLAINTEXT,PLAINTEXT_HOST:PLAINTEXT
      KAFKA_INTER_BROKER_LISTENER_NAME: PLAINTEXT
      KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR: 1

And also you have to install kafka-python-package

pip install kafka-python

All things done! Let's get into coding

Summary

Let's think from frontend, suppose user will request of buying ticket from fronted, behind the scene our kafka producer will send streams of data to the Kafka cluster with respected data. And one kafka consumer will allows applications to read streams of data from the cluster. And we can make as producer and consumer API as much as we need for different different task.

Code

producer.py

import json
from faker import Faker
# faker package just use for some random data

fake = Faker() 
from kafka import KafkaProducer

producer = KafkaProducer(bootstrap_servers="localhost:29092")
for i in range(40_000):

    data = {
        "tiket_id":i,
        "user_id": fake.name(),
        "price": 100,
        "bank_account": fake.bban()
    }

    producer.send("order_details", json.dumps(data).encode("utf-8"))
# order details: is just an event name in kafka cluster
    print(f"done sending ..{i}")

transaction.py responsible for receiving those data and build another producer in kafka cluster just for calculation total revenue and total amount of sold ticket

from kafka import KafkaConsumer, KafkaProducer
import json

KAFKA_TOPIC = "order_details"

producer = KafkaProducer(bootstrap_servers="localhost:29092")

consumer = KafkaConsumer(
    KAFKA_TOPIC,
    bootstrap_servers="localhost:29092"
)

print('start listening"')

while True:
    for i in consumer:
        print('ongoing transaction')
        consumed_message = json.loads(i.value.decode())
        data = {
            "price":consumed_message["price"] #retrieve price data from previous producer which is now in kafka cluster 
        }

        producer.send("analytics", json.dumps(data).encode("utf-8")) #then just demo purpose I created another producer for calculation
        print("Successful transaction..")

analytics.py

from kafka import KafkaConsumer, KafkaProducer
import json

KAFKA_TOPIC = "analytics"
consumer = KafkaConsumer(
    KAFKA_TOPIC,
    bootstrap_servers="localhost:29092"
)
print('start listening"')

while True:
    total_ticket_sell = 0
    revenue = 0
    for i in consumer:
        consumed_message = json.loads(i.value.decode())
        total_ticket_sell += 1
        revenue += consumed_message['price']
        print("============\n\n")
        print("Receiving order")
        print('----------------')
        print(f'total ticket sell so far: {total_ticket_sell}')
        print(f'total revenue so far: {revenue}')

Done, now time for testing. Keep in mind you run your docker compose file and 3 python file in 3 different terminal just to see what happening.

N:B run 2 consumer file at first, then the producer.py file