systemdrill.
THE ENGINEER’S FIELD GUIDESenior / Staff edition
LESS MEMORIZATION. MORE REASONING.

Learn the problem.
Then design the system.

Go beyond “design the next big app.” Understand the invariants,
failure modes, and trade-offs that shape every good architecture.

Open the study bookHow to use this book ↗
THINK IN FIRST PRINCIPLES
◎
01 / PROTECTThe invariant
↓what must remain true?
ϟ
02 / CHALLENGEThe failure
↓where does simple break?
◇
03 / REASONThe pattern → trade-off
13problem families
260recall questions
8comparison sheets
One principle: every component
must earn its place.
THE CORE CURRICULUM

Same building blocks. Different hard problems.

Pick a family. Find its invariant. Follow the reasoning.

01 — 13
CHAPTER 01

Read-heavy / Key-Value Systems

Fast lookups are easy. Correct mappings, hot keys, and stale reads are the real design problem.

ScaleCaching
12 min read20 recall questions
CHAPTER 02

Social Feed / Timeline Systems

Balance fan-out cost, fresh ranking, stable pagination, and the celebrity problem.

ScaleFan-out
11 min read20 recall questions
CHAPTER 03

Messaging / Chat Systems

Separate durable acceptance from live delivery, ordering, and multi-device synchronization.

Real timeOrdering
11 min read20 recall questions
CHAPTER 04

Booking / Reservation Systems

Protect single ownership under contention, expiring holds, and uncertain payment outcomes.

CorrectnessConcurrency
13 min read20 recall questions
CHAPTER 05

Payments / Wallet / Ledger Systems

Model money as auditable entries and treat uncertain external outcomes as first-class states.

CorrectnessIdempotency
12 min read20 recall questions
CHAPTER 06

E-commerce / Inventory / Orders

Coordinate stock, payment, and fulfillment without pretending they share one transaction.

CorrectnessSagas
10 min read20 recall questions
CHAPTER 07

Geospatial / Dispatch Systems

Use approximate location to find candidates, then make assignment authoritative.

Real timeSpatial indexing
11 min read20 recall questions
CHAPTER 08

File Storage / Synchronization

Keep file versions atomic while uploads, offline edits, and garbage collection race.

CorrectnessVersioning
11 min read20 recall questions
CHAPTER 09

Collaborative Editing

Preserve concurrent intent while replicas diverge, reconnect, and converge.

Real timeConvergence
11 min read20 recall questions
CHAPTER 10

Media Streaming / Delivery

Separate reliable publication from adaptive playback, expensive processing, and live latency.

ScaleCDN
10 min read20 recall questions
CHAPTER 11

Search / Indexing / Crawling

Build a recoverable derived index while managing freshness, ranking, and crawl pressure.

ScaleIndexing
10 min read20 recall questions
CHAPTER 12

Notifications / Webhooks / API Infrastructure

Deliver recoverable work without retry storms, duplicate effects, or noisy-neighbor collapse.

ScaleRetries
11 min read20 recall questions
CHAPTER 13

Analytics / Counters / Stream Processing

Define what a number means before choosing how to count, window, replay, or approximate it.

ScaleEvent time
11 min read20 recall questions

13 guides · Designed to be studied, questioned, and revisited.

Connect the patterns

See where the same mechanism solves a different problem.

Find your next question

Decision trees for the moments when “it depends.”

systemdrill.

Understand the constraint. Explain the choice.

A living, Markdown-first study book