🛒 Full Stack Microservices E-Commerce Platform

A production-grade full-stack e-commerce platform built using Spring Boot microservices with API Gateway, Eureka service discovery, Kafka event-driven communication, Saga pattern for distributed transactions, Redis caching, Elasticsearch search, and Razorpay payment integration.

---

🏗 System Architecture Overview

This diagram represents the complete microservices ecosystem including:

• API Gateway for routing
• Eureka for service discovery
• Kafka for asynchronous event communication
• Redis for cart storage & idempotency
• Elasticsearch for search functionality
• Separate database per service
• Razorpay external payment integration

---

🖥 Frontend

The frontend application provides:

• Product browsing
• Search functionality
• Cart management
• Checkout flow
• Order history
• User authentication

The frontend communicates with backend services through the API Gateway.

---

🔹 Microservices Overview

Core Commerce Services

• product-service – Manages product catalog.
• search-service – Handles search functionality using Elasticsearch.
• cart-service – Stores cart data in Redis (key-value store).
• checkout-service – Manages checkout session & idempotency.
• order-service – Handles order lifecycle & acts as Saga Orchestrator.
• inventory-service – Manages stock reservation & release.
• payment-service – Integrates with Razorpay and manages payments.

User Domain Services

• user-service – User profile & authentication.
• wishlist-service – Wishlist management.
• rating-service – Product ratings & reviews.

Supporting Services

• notification-service – Handles notifications.

---

🔄 Order Creation Flow

Flow Summary

1. Checkout-service calls order-service (sync REST).
2. Order-service creates order (status = PENDING).
3. Order-service locks inventory (sync REST).
4. Order-service creates payment (sync REST).
5. Razorpay webhook confirms payment.
6. Payment-service publishes PaymentSuccessEvent to Kafka.
7. Order-service consumes event and updates status → PLACED.

---

🔁 Cancellation Saga Flow (Orchestrated Saga)

Compensation Steps

1. Order status → CANCEL_REQUESTED.
2. Order-service publishes OrderCancelEvent.
3. Inventory-service releases stock.
4. Payment-service processes refund.
5. Saga state tracked in saga_state table.
6. When both compensation steps complete → status = CANCELLED.

---

📦 Transactional Outbox Pattern

To ensure reliable event publishing:

• inventory-service uses outbox_event table.
• order-service uses order_outbox table.
• Events are written in the same DB transaction.
• Scheduled publisher sends events to Kafka.
• Prevents data inconsistency and event loss.

---

🔁 Communication Model

Synchronous (REST / Feign)

• checkout → order
• order → inventory
• order → payment

Asynchronous (Kafka Events)

• PaymentSuccessEvent
• PaymentRefundedEvent
• InventoryReleasedEvent
• OrderCancelEvent

---

🗄 Storage & Data Layer

• Separate database per microservice.
• Redis used by:
  • cart-service (cart storage)
  • checkout-service (idempotency)
• Elasticsearch used by search-service.
• Saga state stored in saga_state.
• Outbox tables used for reliable messaging.

---

🛠 Tech Stack

Backend

• Java
• Spring Boot
• Spring Data JPA
• Spring Kafka
• Feign Client
• Resilience4j

Frontend

• (Add your frontend tech here)

Infrastructure

• Kafka
• Redis
• Elasticsearch
• MySQL / PostgreSQL
• Docker
• Maven

---

🚀 Running the Project Locally

Prerequisites

• Docker
• Kafka
• Redis
• MySQL / PostgreSQL

Start Backend

docker-compose up --build

Start Frontend

npm install
npm start

---

🔐 Environment Configuration

Configure the following environment variables:

• Razorpay API Keys
• Kafka Broker URL
• Database credentials
• Redis configuration

---

🛡 Resilience & Reliability

• Circuit breaker using Resilience4j
• Retry mechanisms
• Saga retry scheduler
• Idempotent consumers
• Transactional outbox pattern
• At-least-once Kafka delivery

---

⚠️ Troubleshooting

Common issues:

• Kafka consumer not receiving messages
• Razorpay webhook signature validation failure
• Duplicate event processing
• Saga stuck in intermediate state
• Redis cache expiration issues