Plate I · notes on System Design Interview, Vol 1 and 2

Read the notes, then the papers behind them.

Chapter-by-chapter notes on the System Design Interview books — twenty-eight designs, from scaling to a stock exchange. Nine chapters point past the book to primary sources: Amazon’s Dynamo paper, Google’s BigTable and Maglev, Twitter’s Snowflake and more.

Read the notes on GitHub The books

01 scale from zero → 28 stock exchange, in reading ordernote lengths read from the repo · Oct 8, 2026

01020304050607080910111213141516171819202122232425262728longest notes · 28.6 KB

star size = length of the notes, 3.5 to 28.6 KB outside reading listed in the README one source

What the chart shows

Every chapter with outside reading sits in the first fifteen: thirty sources across nine chapters.

The longest notes — chapters 19, 28 and 21 — list none yet. The README calls the notes a work in progress.

04 · 2Rate Limiter

Circuit Breaker Algorithm · Uber Rate Limiter

05 · 5Consistent Hashing

Consistent Hashing · Stanford CS168 · Apache Cassandra · Scaling Discord · Google Maglev

06 · 6Key-Value Store

Amazon Dynamo · Cassandra Architecture · Google BigTable · Dynamo DB internals, design patterns and two talks

07 · 2Unique ID Generator

Ticket Servers: Distributed Unique Primary Keys on the Cheap · Snowflake

09 · 2Web Crawler

Web Crawling · Google Dynamic Rendering

12 · 2Chat System

How Discord stores billions of messages · Flannel, Slack’s edge cache

13 · 2Search Autocomplete

How We Built Prefixy · Prefix Hash Tree

14 · 6YouTube

YouTube Architecture · YouTube scalability 2012 · transcoding at scale · Facebook live video · two Netflix encoding posts

15 · 3Google Drive

Differential Synchronization, paper and talk · How We’ve Scaled Dropbox