Top 30 System Design Concepts

29 Mar 2025

Client-Server Architecture: The Backbone of Modern Web Applications

• Ever wondered how your favorite apps and websites actually work behind the scenes?
• The fascinating world of client-server architecture
• The foundation that powers virtually all modern web and mobile applications.
• Client sends request to the server over the internet, server response back with the data from the database, the is rendered in the screen by the client
• This 3 tier architecture is used in most modern web and mobile application

IP Address: The Digital Street Address

• Every server deployed on the public internet has its own unique IP address
• Clients use these IP addresses to communicate with servers over the internet
• Think of IP addresses as the postal addresses of the digital world—without them, your data wouldn’t know where to go!

Domain Names: Because Who Remembers Numbers?

• IP addresses (like 172.217.168.238) are about as memorable as your tax ID number
• Domain names (like google.com) provide a human-friendly alternative
• When you type a domain name, your Internet Service Provider (ISP) connects to a Domain Name System (DNS) to resolve it to the correct IP address
• It’s like having a contact list in your phone instead of memorizing everyone’s number!

Proxy & Reverse Proxy: The Digital Middlemen

Proxy

• Acts as a middleman between you and the internet
• Makes requests on your behalf, hiding your IP address from the websites you visit
• Like having a friend buy concert tickets for you so the venue doesn’t know who’s coming

Reverse Proxy

• Works in the opposite direction, intercepting client requests before they reach the server
• Forwards requests to the appropriate backend servers
• Think of it as a receptionist who directs visitors to the right department

Latency: The Digital Distance Tax

• Latency is the delay caused by the physical distance between client and server
• The farther away the server, the longer it takes for data to travel
• One solution? Deploy multiple servers across the world
• Because nobody likes waiting for a page to load while their coffee gets cold!

HTTP/HTTPS: The Communication Protocol

• HTTP (Hypertext Transfer Protocol) is how clients and servers talk to each other
• HTTPS is the secure, encrypted version of HTTP
• Without HTTPS, your data travels naked across the internet—not a good look!

API: The Digital Translator

• APIs (Application Programming Interfaces) act as intermediaries between clients and servers
• Clients send requests via HTTP to the API in a specific format
• The API processes these requests and returns responses
• It’s like a waiter taking your order, delivering it to the kitchen, and bringing back your food

REST: Rules of Engagement

• REST (Representational State Transfer) defines a structured way for clients and servers to communicate
• Key characteristics:
  • Stateless: Every request is independent (the server has no memory of previous requests)
  • Resource-oriented: Everything is treated as a resource
  • Standard methods: GET, POST, PUT, DELETE
• It’s like having a standardized language so everyone can understand each other

GraphQL: The Picky Eater

• REST can be inefficient, often returning unwanted data or requiring multiple requests
• GraphQL lets you specify exactly what data you need—no more, no less
• Comes with trade-offs: requires more backend processing and is harder to cache
• Like ordering a custom sandwich instead of taking whatever’s on the menu

Databases: The Digital Filing Cabinet

• Databases ensure data is stored, retrieved, and managed efficiently
• They keep your data secure, consistent, and durable
• Without them, your favorite apps would have amnesia!

SQL vs NoSQL: Different Strokes for Different Folks

SQL (Relational Databases)

• Uses tables with rows and columns
• Follows ACID properties (Atomicity, Consistency, Isolation, Durability)
• Provides strong consistency
• Perfect for structured data with clear relationships

NoSQL

• Comes in various flavors: key-value stores, document stores, graph databases, wide-column stores
• Offers high scalability and performance
• Features flexible schemas
• Ideal for unstructured or semi-structured data
• When your data doesn’t fit neatly into tables, NoSQL is your friend

Scaling: Growing Pains

Vertical Scaling

• Adding more resources (storage, compute, RAM) to a single machine
• Quick fix but not a long-term solution
• Every machine has its limits
• Like trying to make your car faster by adding a bigger engine—eventually, you hit physical constraints

Horizontal Scaling

• Distributing requests among multiple servers
• Reduces demand on any single system
• Eliminates single points of failure
• Like adding more lanes to a highway instead of making one lane wider

Load Balancer: The Traffic Director

• When scaling horizontally, how does a client know which server to talk to?
• Load balancers take in all requests and distribute them to the appropriate servers
• Distribution can be based on various strategies (round-robin, least connections, etc.)
• They also collect responses from servers and return them to clients
• Think of them as air traffic controllers for your data

Indexing: The Digital Shortcut

• As data grows, retrieval and updates become slower
• Indexing creates efficient lookup tables to find values without scanning entire tables
• Like the index in a book—it helps you find specific topics quickly
• Speeds up reads but slows down writes (since indexes need updating)
• No free lunch in computer science!

Replication: Copy That

• When requests are very high, indexing alone isn’t enough
• Creating copies of the database allows load distribution
• Write operations happen on the primary replica and get copied to others
• Improves availability and is great for read-heavy databases
• It’s like having multiple copies of a popular book in a library

Sharding: Divide and Conquer

• When data grows to terabytes, even replicated databases struggle
• Sharding splits the main database into multiple shards, each containing a subset of data
• Data is divided based on a sharding key (e.g., primary key)
• Queries are distributed, improving both read and write performance
• Like splitting a massive encyclopedia into multiple volumes

Vertical Partitioning: Column Control

• Sometimes the issue isn’t the number of rows but the number of columns
• Vertical partitioning splits tables into smaller ones (normalization)
• Separates chunks of data that can be split from the primary table
• Like organizing your closet by type of clothing instead of throwing everything together

Caching: The Speed Demon

• Retrieving data from disk is always slower than from memory
• Caching stores frequently accessed data in memory for faster retrieval
• The cache-aside pattern checks the cache first before going to the database
• Time-to-live (TTL) values prevent serving outdated data
• It’s like keeping your most-used cooking ingredients on the counter instead of in the pantry

Denormalization: Breaking the Rules

• Normalized databases reduce redundancy but can require many joins
• Joins slow down as datasets grow
• Denormalization reduces joins by combining related data into a single table
• Accepts some data duplication for performance gains
• Used primarily in read-heavy applications
• Sometimes you have to break the rules to win the game!

CAP Theorem: Pick Two, Not Three

• No distributed database system can simultaneously guarantee all three:
  • Consistency: All nodes see the same data at the same time
  • Availability: Every request receives a response
  • Partition Tolerance: System continues to function despite network failures
• In real-world scenarios, network partitions are inevitable, so we choose between CP and AP
• CP systems prioritize consistency over availability
• AP systems prioritize availability over consistency (think social media likes, view counts)
• It’s the “good, fast, cheap—pick two” of database design

Blob Storage: The Digital Warehouse

• Services like Amazon S3 store large structured files (images, videos, documents)
• Related data is stored in buckets with unique URLs for easy access
• Advantages: scalability, pay-as-you-go model, automatic replication
• Streaming directly from blob storage can be slow for distant users
• Great for storing your cat videos, not so great for serving them quickly worldwide

Content Delivery Network (CDN): The Global Express

• CDNs are distributed servers across the world
• They deliver HTML, JavaScript, images, and videos based on geographical location
• Solves the problem of slow loading from distant servers
• Like having local warehouses for Amazon instead of shipping everything from headquarters

Real-time Applications: The Instant Gratification

• Live chat, stock markets, and online games need real-time updates
• Frequent polling is inefficient and increases server load
• WebSockets provide continuous two-way communication between client and server
• Like having a phone call instead of repeatedly asking “Any news?”

Webhooks: The Notification System

• Instead of constantly checking for updates, webhooks notify you when events occur
• Applications register a webhook URL with a server
• When an event occurs, the server sends an HTTP request to that URL
• Like having someone call you when your package arrives instead of checking the mailbox every hour

Microservices: The Modular Approach

• Traditional monolithic applications have all services in one application
• This creates single points of failure and makes code hard to manage
• Microservices split applications into separate servers that scale independently
• They communicate via HTTP or message queues
• Like having specialized teams instead of everyone trying to do everything

Message Queues: The Asynchronous Communicator

• Synchronous communication (HTTP) doesn’t scale well for inter-server communication
• Message queues enable asynchronous communication
• Producers place messages in the queue and forget about them
• Consumers retrieve messages when they’re ready to process them
• Example: Payment service queues notification requests instead of waiting for notifications to send
• It’s like dropping off your dry cleaning and coming back later, instead of waiting there

Rate Limiting: The Bouncer

• Public services need protection from bots and abusive users
• Rate limiting assigns quotas per user/IP address
• Exceeding quotas results in blocked requests
• Common algorithms: fixed window, sliding window, token bucket
• Like a nightclub that only lets in a certain number of people per hour

API Gateway: The Swiss Army Knife

• Handles rate limiting, authentication, logging, monitoring, and request routing
• Acts as a single entry point to your microservices
• Simplifies the client interface to your complex backend
• Like having a concierge desk at a large hotel

Idempotency: The Safety Net

• In distributed networks, service failures and retries are common
• Accidental refreshes during payments could result in double charges
• Idempotency ensures repeated requests produce the same result as a single request
• Each request gets a unique ID; the server checks if it’s already been handled
• Like having a system that recognizes you’ve already paid, even if you accidentally hit “Pay” twice

---

There you have it—the essential components that power the digital experiences we take for granted every day. Next time you’re scrolling through social media or ordering food online, you’ll have a better appreciation for the complex architecture working behind the scenes!