Case files · engineering stories

Case files.

Production incidents, rush hours and the things I built, told as stories with blueprints. Each one ends with the lessons you can steal.

Your record0 of 13 closed · 0 solved ◆

On-call badge: read a case file to the end.

13 case files · newest first

CASE-012LATESTINCIDENT · 18 MIN READ · OCT 16, 2026The $440M Deploy: Knight Capital's 45 MinutesOne server out of eight missed a deploy. A recycled feature flag woke up code retired nine years earlier, and in 45 minutes a Wall Street trading firm lost about $440 million. This is how it happened.DeploymentFeature FlagsDead CodeKill SwitchOpen the case file →+50 XP

CASE-013 · SYSTEM DESIGN · 19 MIN · OCT 202625.3 Million People, One Cricket Match: Why Hotstar Scaled Before the Star Walked InJuly 2019. India are 5 for 3, Dhoni walks in, and a million new viewers arrive every minute. How Hotstar served a world-record 25.3 million at once by scaling before the moment, not after it.System DesignHotstarLive Streaming
CASE-011 · INCIDENT · 19 MIN · OCT 2026One File, 8.5 Million Blue ScreensOn July 19, 2024, a routine security content update reached Windows machines worldwide in 78 minutes and crashed about 8.5 million of them. The cause: a rule that asked for field 21 when the sensor only had 20.Staged RolloutsCanary DeploymentsConfiguration
CASE-010 · SYSTEM DESIGN · 15 MIN · AUG 2026The Tatkal Gamble: How India Sells 500 Train Seats to 2 Million People Without a Single MistakeEvery morning at 10 AM, millions of Indians fight for a few hundred train seats in a battle that lasts less than a minute. This is the story of the system that referees it — fairly, instantly, and without ever selling the same seat twice.System DesignIRCTCTicket Booking
CASE-009 · BUILD LOG · 23 MIN · NOV 2025From Office-Bound Devices to Global Access: The BridgeLink StoryHow I built BridgeLink, an open-source CLI that makes the Android phones on your desk usable from anywhere, and what it could save a team versus cloud device farms.Android DevelopmentRemote DevelopmentADB
CASE-008 · BUILD LOG · 16 MIN · AUG 2025From Mac Bottleneck to Universal Access: The iOS Bridge StoryHow I built iOS Bridge, an open-source tool that streams Mac-hosted iOS simulators to Windows, Linux and the browser, and what one shared Mac could save a team.iOS DevelopmentCross-PlatformWebRTC
CASE-007 · SYSTEM DESIGN · 25 MIN · AUG 2025Building a Scalable Spam Detection System for User-Submitted URLs Using Bloom FiltersHow to design a low-latency, large-scale spam URL detection service using probabilistic data structures.Bloom FilterDistributed SystemsURL Filtering
CASE-006 · SYSTEM DESIGN · 17 MIN · AUG 2025Building DIGIPIN: From Chaos to 4-Meter Precision Across India's Digital GeographyHow to design a nationwide location-to-code service that assigns unique alphanumeric addresses to every 4m×4m grid in India.Geospatial SystemsGeoHashLocation Services
CASE-005 · INCIDENT · 12 MIN · AUG 2025The Great ID Crisis: When Auto-Increment Hits the Mathematical WallA real-world incident story of integer overflow bringing down a high-volume analytics system and the emergency response strategies that saved the day.Integer OverflowDatabase DesignAuto-increment
CASE-004 · INCIDENT · 17 MIN · AUG 2025The Silent Cache Poisoning Disaster: When CDN Becomes Your EnemyA Black Friday nightmare where a misconfigured CDN turned private customer data into a public cache, exposing shopping carts and personal information across user sessions.CDNCache PoisoningSecurity
CASE-003 · INCIDENT · 8 MIN · AUG 2025Handling Node Outages in Leader-Follower Database Replication: A Tale of ResilienceHow distributed database systems maintain availability and consistency when nodes fail in leader-follower architectures.Database ReplicationDistributed SystemsHigh Availability
CASE-002 · SYSTEM DESIGN · 18 MIN · AUG 2025The Great Video Call Revolution: When WebSockets Met WebRTC and Changed EverythingA tale of two technologies fighting for supremacy in real-time video communication, and why choosing the wrong one can crash your startup dreams.WebSocketsWebRTCVideo Streaming
CASE-001 · INCIDENT · 27 MIN · AUG 2025The Great Travel Booking Disaster: How 2PC Nearly Killed Our Business and Saga Pattern Saved ItA tale of distributed transactions gone wrong and the architectural patterns that brought resilience to global travel bookings.Distributed TransactionsTwo Phase CommitSaga Pattern

Get the next case file by email

One story when it ships. No digest, no spam.