🚀 Sales Kafka - Orchestration Saga Pattern

Monorepo implementing the Saga Pattern (Orchestration) using Java, Spring Boot, Apache Kafka, PostgreSQL, MongoDB, and Docker.

📑 Table of Contents

• 🛠️ Technologies
• 🏗️ Architecture
• 🗺️ Topic Mapping
• ☕ Getting Started
• 🔀 Choreography Version

🛠️ Technologies

• Java 25
• Spring Boot 4.1.0
• Gradle 9.6
• Apache Kafka
• API REST
• PostgreSQL
• MongoDB
• Docker
• Redpanda Console

🏗️ Architecture

🎼 Orchestrated Saga Architecture

• Order-Service: microservice responsible only for generating an initial request and receiving a notification. Here we have REST endpoints to start the process and retrieve event data. The database used will be MongoDB.
• Orchestrator-Service: microservice responsible for orchestrating the entire Saga execution flow, it will know which microservice was executed and in which state, and which will be the next microservice to be sent, this microservice will also save the process from events. This service does not have a database.
• Product-Validation-Service: microservice responsible for validating whether the product specified in the order exists and is valid. This microservice will store a product validation for an order ID. The database used will be PostgreSQL.
• Payment-Service: microservice responsible for making a payment based on the unit values ​​and quantities informed in the order. This microservice will store the payment information for an order. The database used will be PostgreSQL.
• Inventory-Service: microservice responsible for lowering the stock of products from an order. This microservice will store the download information of a product for an order ID. The database used will be PostgreSQL.

All services will go up through docker-compose.yml.

🗺️ Topic Mapping

Event	Service	Kafka Topic	Type
PRODUCT_VALIDATION_SUCCESS	order-service	notify-ending	consumer
FINISH_FAIL	order-service	start-saga	producer
PRODUCT_VALIDATION_FAIL	orchestrator	orchestrator	consumer
FINISH_FAIL	orchestrator	finish-success	producer/consumer
PAYMENT_SUCCESS	orchestrator	finish-fail	producer/consumer
PAYMENT_FAIL	orchestrator	notify-ending	producer
PRODUCT_VALIDATION_FAIL	product-validation-service	product-validation-success	consumer
INVENTORY_SUCCESS	product-validation-service	product-validation-fail	consumer
INVENTORY_FAIL	product-validation-service	orchestrator	producer
PAYMENT_FAIL	payment-service	payment-success	consumer
FINISH_SUCCESS	payment-service	payment-fail	consumer
	payment-service	orchestrator	producer
	inventory-service	inventory-success	consumer
	inventory-service	inventory-fail	consumer
	inventory-service	orchestrator	producer

▶️ Getting Started

☕ How to run

There are several ways to run:

🐳 Option 1 — Docker Compose

Running everything via docker-compose

⚙️ Option 2 — Build Script

Running everything through the automation script that i made available (build-bash.sh)

💻 Option 3 — Run Infrastructure Only

Running only the database and message broker (Kafka) services separately

☕ Option 4 — Run Services Manually

Running applications manually via CLI (java -jar or gradle bootRun or via IntelliJ)

📦 Requirements

To run the applications, you will need to have installed:

• Docker
• Java 25
• Gradle 9.6

🔀 Choreography Version

👉 Check out the choreography-based implementation: https://github.com/Stephulz/sales-kafka-saga-choreographed