In modern e-commerce systems (like Amazon, Flipkart), when a user buys an item and it gets delivered, many microservices work together.

These services communicate using a mix of synchronous and asynchronous interactions to make the system fast, scalable, and reliable.

Let's walk through the full journey! 🚀

🛒 Step 1: User Selects an Item and Places Order

🔥 Services involved:

• Product Catalog Service
• Cart Service
• Order Service
• User Service

📞 Communication:

• User App → Product Catalog Service: (Sync, REST API) to fetch item list.
• User App → Cart Service: (Sync, REST API) to add item to cart.
• User App → Order Service: (Sync, REST API) to create an order.

✅ Why synchronous here?

• User expects instant response: "Show me the items", "Add to cart", "Order placed successfully."
• Immediate success/failure feedback.

💳 Step 2: Payment Processing

🔥 Services involved:

• Payment Service
• Order Service

📞 Communication:

• Order Service → Payment Service: (Sync, REST API or gRPC) to process the payment.

✅ Why synchronous here?

• Payment must succeed or fail instantly.
• User must see "Payment Successful" or "Payment Failed" immediately.

📬 Step 3: Order Confirmation & Inventory Management

🔥 Services involved:

• Order Service
• Inventory Service
• Notification Service

📩 Communication:

• Order Service → Inventory Service: (Async, via Kafka event ORDER_CONFIRMED) to reduce stock.
• Order Service → Notification Service: (Async, via Kafka event ORDER_PLACED) to send email/SMS.

✅ Why asynchronous here?

• Reducing stock or sending email is background work.
• It shouldn't delay the "Order placed" response to the user.
• If inventory update or notification fails, it can be retried separately.

🚚 Step 4: Delivery Arrangement

🔥 Services involved:

• Delivery Service
• Order Service
• Inventory Service

📩 Communication:

• Inventory Service → Delivery Service: (Async, Kafka event like STOCK_RESERVED) to schedule pickup.
• Delivery Service → Order Service: (Async, Kafka event like DELIVERY_ASSIGNED) to update delivery status.

✅ Why asynchronous?

• Delivery scheduling can take some time.
• If assigning a delivery boy or a delivery slot takes 2-5 seconds, the user shouldn’t be blocked waiting.

📦 Step 5: Shipment and Delivery Tracking

🔥 Services involved:

• Delivery Service
• Tracking Service
• Notification Service

📞 Communication:

• User App → Tracking Service: (Sync, REST API) to get live delivery updates.

• Delivery Service → Tracking Service: (Async, Kafka event DELIVERY_STARTED, OUT_FOR_DELIVERY).

• Tracking Service → Notification Service: (Async, Kafka event DELIVERY_UPDATE) to send SMS/email.

✅ Why mixed?

• User polls tracking info synchronously for real-time updates.
• Internal updates like "out for delivery" are sent asynchronously.

🧩 Full Service Map (with communication type)

🎯 Why do we use Sync vs Async?

🔥 Key Takeaways

• Use Sync when user interaction is happening.
• Use Async when background services communicate.
• Kafka, RabbitMQ, or EventBridge are used for async.
• REST APIs or gRPC are used for synchronous needs.
• Scalability and resiliency are achieved by decoupling services via events.

📚 Real-World Example:

• Amazon uses a heavy combination of synchronous APIs and asynchronous event-driven microservices.
• Payments, Search, Cart = Mostly synchronous.
• Inventory, Delivery, Notification, Fraud Check = Mostly asynchronous via events.

✍️ Final Thought

Building a robust e-commerce platform involves understanding where to synchronize and where to decouple with asynchronous messaging.

Balancing these two wisely is the real art of microservice system design!