Order tracking state machine in LLD - System Design Exercise

Design: Order Tracking State Machine

Design the state machine logic, API interface, and data model for order states. Out of scope: payment processing, inventory management.

Functional Requirements

FR1: Track the state of an order from placement to delivery

FR2: Support states: Created, Confirmed, Packed, Shipped, Out for Delivery, Delivered, Cancelled

FR3: Allow valid transitions only (e.g., cannot ship before packing)

FR4: Notify users on state changes

FR5: Handle concurrent updates safely

FR6: Provide API to query current order state

FR7: Support at least 10,000 orders concurrently

Non-Functional Requirements

NFR1: API response latency p99 < 100ms

NFR2: System availability 99.9%

NFR3: State transitions must be consistent and atomic

NFR4: Scalable to handle peak loads during sales events

Think Before You Design

Questions to Ask

❓ Question 1

❓ Question 2

❓ Question 3

❓ Question 4

❓ Question 5

❓ Question 6

Key Components

State machine engine or library

Order state database

API server for client interaction

Notification service

Concurrency control mechanism

Design Patterns

State machine pattern

Event sourcing for state changes

Observer pattern for notifications

Optimistic locking for concurrency

RESTful API design

Reference Architecture

API Server

↓

Notification Service

Components

API Server

Node.js/Express or Python/Flask

Expose REST API endpoints for querying and updating order states

State Machine Logic

Custom code or state machine library

Enforce valid state transitions and business rules

Order State Database

Relational DB (PostgreSQL) or NoSQL (DynamoDB)

Persist order states and transition history

Notification Service

Message queue (RabbitMQ) + Email/SMS gateway

Send asynchronous notifications on state changes

Concurrency Control

Optimistic locking or transactions

Prevent conflicting state updates

Request Flow

1. Client sends API request to update order state

2. API Server receives request and forwards to State Machine Logic

3. State Machine Logic validates requested transition against current state

4. If valid, State Machine Logic updates Order State Database atomically

5. State Machine Logic triggers Notification Service asynchronously

6. API Server responds with updated order state

7. Client can query current order state via API

Database Schema

Entities: - Order: order_id (PK), current_state, user_id, timestamps - OrderStateHistory: id (PK), order_id (FK), previous_state, new_state, changed_at Relationships: - One Order has many OrderStateHistory records - current_state in Order reflects latest state - State transitions validated before updating current_state

Scaling Discussion

Bottlenecks

Database write contention on popular orders during peak times

Notification service overload with many state changes

API server CPU/memory limits under high concurrent requests

State machine logic becoming a single point of failure

Solutions

Use database sharding or partitioning by order_id to distribute load

Implement batching and rate limiting in Notification Service

Deploy API servers behind load balancer with autoscaling

Design State Machine Logic as stateless microservice instances with shared DB

Interview Tips

Time: 10 min for requirements and clarifications, 15 min for architecture and data model, 10 min for scaling and trade-offs, 10 min for Q&A

Clearly define states and allowed transitions

Explain concurrency control to maintain consistency

Discuss asynchronous notifications for better performance

Justify technology choices based on requirements

Highlight scalability strategies for peak loads