A kafka spring boot microservice application

This repo contains two types of spring boot applications:

• consumer- and producer services which can be run together
• and a multimodule spring cloud stream application.

For the spring kafka producer-consumer application, we demonstrate how a loose coupling can be implemented
between services (producer- and consumer-service) by using kafka. The producer-service fires an event which is consumed by the consumer-service.

The second part of the repository consists of a multimodule spring boot microservice application (util-service, user-service and user-composite). The multi-module application demonstrates how to implement a loosely coupled rabbitMQ-Kafka, distributed messaging application developed using the cloud streams functional API a (via the spring cloud framework).

Tech Stack

• kafka and spring cloud (kafka streams)
• spring-boot v.3.1+
• mysql
• spring cloud
• spring JPA (for persistence, etc)
• Spring web
• docker and docker-compose
• jdk 17+
• maven
• rabbitmq
• kafka

Steps to run

Clone the repository using:

git clone <name of repo>

And then change to the directory

cd spring-kafka-examples

For the purpose of this tutorial (or dev), I ran kafka via docker-compose using

docker-compose up -d

To run the application, use

mvn spring-boot:run

To install kafka, you can use the instructions provided on ```https://kafka.apache.org/quickstart``, or use an enterprise based kafka. or install kafka using the instructions on the Kafka homepage.

Contents

The directories consist of the repository consist of the following projects:

• Kafka producer service named producer-service to produce messages
• And a consumer service named consumer-service to consumer messages

Composing Services

• A composite service named collect services provides composite implementation to the ùser- and util services
• the utils service contains utility classes and methods, while the
• the user-service contains basic CRUD operations for the user-service.

Notes

When a call is made such as to create a user (via the corresponding endpoint provided in the collect-services), an event is registered in kafka/rabbitMQ. The collect-servicescalls the ùser-service by using a distributed architecture (via kafka). By registering the event using the choices defined in the util-service, an event bus alerts other services which can serve as consumers. This design pattern takes care of situations whereby database updates can be performed in an optimal pattern between microservices. It also enforces the notion of using loose coupling between participating services, and can be easily extended to include spring Application Events. This method also enables the use of SAGA pattern which is vital in establishing the SAGA orchestration pattern in microservices.

Both the

• Consumer - and
• producer services can be run together.

While the collect-services can be run together with user-, util-, consumer-, producer services. Also note that jdk 17+ (and higher versions) are required to run the examples.