Inventory management in LLD - System Design Exercise

Design: Inventory Management System

Includes product and warehouse management, stock tracking, alerts, and query APIs. Excludes order processing, payment, and shipping logistics.

Functional Requirements

FR1: Track stock levels for multiple products across multiple warehouses

FR2: Support adding new products and warehouses

FR3: Update stock quantities on product purchase, return, or restock

FR4: Provide real-time stock availability queries

FR5: Generate alerts when stock for any product falls below a threshold

FR6: Support concurrent updates without data loss or inconsistency

Non-Functional Requirements

NFR1: Handle up to 10,000 concurrent users querying stock

NFR2: API response time p99 under 200ms for stock queries

NFR3: System availability 99.9% uptime

NFR4: Data consistency for stock updates must be strong to avoid overselling

Think Before You Design

Questions to Ask

❓ Question 1

❓ Question 2

❓ Question 3

❓ Question 4

❓ Question 5

Key Components

API Gateway for client requests

Authentication and Authorization service

Inventory Service to handle stock logic

Database for storing product, warehouse, and stock data

Cache layer for fast stock queries

Alerting service for low stock notifications

Message queue for handling asynchronous updates

Design Patterns

CQRS (Command Query Responsibility Segregation) for separating reads and writes

Event sourcing for tracking stock changes

Caching strategies to reduce database load

Distributed locking or transactions for concurrency control

Reference Architecture

Client
  |
  v
API Gateway
  |
  v
Authentication Service
  |
  v
Inventory Service <--> Cache Layer
  |
  v
Database
  |
  v
Alerting Service
  
Message Queue (for async stock updates and alerts)

Components

API Gateway

Nginx or AWS API Gateway

Handle client requests, route to services, enforce rate limiting

Authentication Service

OAuth 2.0 / JWT

Authenticate users and authorize access to inventory APIs

Inventory Service

Node.js / Python microservice

Process stock updates, queries, and business logic

Database

PostgreSQL

Store product, warehouse, and stock data with strong consistency

Cache Layer

Redis

Cache stock levels for fast read access

Alerting Service

Custom microservice or AWS SNS

Send notifications when stock falls below threshold

Message Queue

RabbitMQ or AWS SQS

Handle asynchronous stock update events and alert triggers

Request Flow

1. Client sends stock query or update request to API Gateway

2. API Gateway forwards request to Authentication Service for user validation

3. Authenticated requests reach Inventory Service

4. For stock queries, Inventory Service first checks Cache Layer

5. If cache miss, Inventory Service queries Database and updates cache

6. For stock updates, Inventory Service writes to Database within transaction

7. Inventory Service publishes stock update event to Message Queue

8. Alerting Service consumes events from Message Queue to check thresholds

9. Alerting Service sends notifications if stock is low

Database Schema

Entities: - Product (product_id PK, name, description, category) - Warehouse (warehouse_id PK, location, name) - Stock (stock_id PK, product_id FK, warehouse_id FK, quantity, last_updated) Relationships: - Product to Stock: 1 to many - Warehouse to Stock: 1 to many Stock table stores current quantity per product per warehouse.

Scaling Discussion

Bottlenecks

Database write contention on stock updates

Cache invalidation and consistency challenges

Message queue overload with high update volume

API Gateway rate limiting under heavy load

Solutions

Use database sharding by warehouse or product to distribute writes

Implement cache with TTL and write-through or write-back strategies

Scale message queue horizontally and partition topics

Deploy multiple API Gateway instances with load balancing

Interview Tips

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

Clarify consistency needs and scale targets early

Explain choice of database for strong consistency

Describe caching strategy to meet latency goals

Discuss concurrency control to prevent overselling

Highlight asynchronous alerting for scalability

Mention how to handle failures and retries

Practice

1. What is the primary purpose of an inventory management system?

easy

A. To track product quantities and prevent stock issues

B. To design product packaging

C. To manage employee schedules

D. To create marketing campaigns

Solution

1. Step 1: Understand inventory management goals

   Inventory management focuses on tracking product quantities to avoid running out or overstocking.

2. Step 2: Eliminate unrelated options

   Options about packaging, schedules, and marketing do not relate to inventory tracking.

3. Final Answer:

   To track product quantities and prevent stock issues -> Option A

4. Quick Check:

   Inventory management = tracking stock [OK]

Hint: Inventory systems track stock levels, not unrelated tasks [OK]

Common Mistakes:

• Confusing inventory with marketing or HR tasks
• Thinking inventory manages packaging design
• Assuming inventory handles employee schedules

2. Which of the following is the correct way to check if an item exists in an inventory dictionary named stock in Python?

easy

A. if stock.has_key('item'):

B. if 'item' in stock:

C. if stock.contains('item'):

D. if stock.exists('item'):

Solution

1. Step 1: Recall Python dictionary syntax

   To check if a key exists in a dictionary, use the in keyword.

2. Step 2: Identify correct syntax

   stock.has_key() is deprecated, and contains or exists are invalid methods.

3. Final Answer:

   if 'item' in stock: -> Option B

4. Quick Check:

   Use 'in' to check keys in dict [OK]

Hint: Use 'in' keyword to check keys in Python dicts [OK]

Common Mistakes:

• Using deprecated has_key() method
• Using non-existent methods like contains()
• Confusing method names for key checks

3. Given the Python code below, what will be the output?

stock = {'apple': 10, 'banana': 5}
stock['apple'] -= 3
print(stock['apple'])

medium

A. Error

B. 13

C. -3

D. 7

Solution

1. Step 1: Understand the initial stock

   Initially, 'apple' has quantity 10.

2. Step 2: Apply the subtraction operation

   Subtracting 3 from 10 results in 7.

3. Final Answer:

   7 -> Option D

4. Quick Check:

   10 - 3 = 7 [OK]

Hint: Subtract quantity correctly to find updated stock [OK]

Common Mistakes:

• Adding instead of subtracting
• Confusing keys or values
• Expecting an error due to subtraction

4. Identify the error in the following inventory update code snippet:

stock = {'apple': 5}
stock['banana'] -= 2
print(stock)

medium

A. No error, banana quantity becomes -2

B. SyntaxError due to invalid subtraction

C. KeyError because 'banana' does not exist in stock

D. TypeError because stock is not a list

Solution

1. Step 1: Check if 'banana' key exists

   'banana' is not in the stock dictionary initially.

2. Step 2: Understand dictionary behavior on missing keys

   Subtracting from a missing key causes a KeyError in Python.

3. Final Answer:

   KeyError because 'banana' does not exist in stock -> Option C

4. Quick Check:

   Missing key access = KeyError [OK]

Hint: Accessing missing dict keys causes KeyError [OK]

Common Mistakes:

• Assuming missing keys default to zero
• Expecting negative values without initialization
• Confusing error types

5. You are designing an inventory system that must handle multiple warehouses. Which design approach best ensures accurate stock counts across warehouses and prevents overselling?

hard

A. Maintain separate stock counts per warehouse and use transactions to update atomically

B. Keep a single global stock count without warehouse details

C. Update stock counts asynchronously without locking

D. Allow negative stock counts to handle overselling

Solution

1. Step 1: Consider multi-warehouse stock tracking

   Each warehouse should have its own stock count to track inventory accurately.

2. Step 2: Ensure atomic updates to prevent overselling

   Using transactions or locks ensures stock updates are consistent and prevent race conditions.

3. Final Answer:

   Maintain separate stock counts per warehouse and use transactions to update atomically -> Option A

4. Quick Check:

   Atomic updates + per-warehouse stock = accurate inventory [OK]

Hint: Use atomic transactions and per-warehouse counts [OK]

Common Mistakes:

• Using global stock ignores warehouse differences
• Updating asynchronously causes race conditions
• Allowing negative stock hides overselling problems