Skip to content

README.md

Latest commit

 

History

History
88 lines (72 loc) · 5.2 KB

README.md

File metadata and controls

88 lines (72 loc) · 5.2 KB
Raw

System Design

Interview-focused system design — one folder per product, each with full architecture, mermaid diagrams, database schemas, indexing strategies, sharding, APIs, and Q&A.

Companies & Products

Folder Product Key challenge What's inside
uber/ Uber Real-time driver matching, geospatial indexing Architecture, sequence, ER, sharding, state machine diagrams · PostgreSQL DDL · Redis/Kafka indexing · Geohash · Q&A
instagram/ Instagram Feed fan-out, media CDN, billion-scale likes Push/pull fan-out diagrams · Cassandra/PostgreSQL schemas · CDN pipeline · Like counter · Q&A
facebook/ Facebook Social graph (TAO), ML feed ranking TAO graph layer · News feed ranking pipeline · Cassandra/MySQL schemas · Graph indexing · Q&A
airbnb/ Airbnb Geo search, booking consistency, payments Search + booking flow diagrams · Elasticsearch geo indexing · PostgreSQL DDL · Payment idempotency · Q&A
url-shortener/ bit.ly / TinyURL Base62 encoding, redirect at scale Create/redirect sequences · DynamoDB + ClickHouse schemas · 3-tier cache · Base62 encoding · Q&A
realtime-coding/ CoderPad / HackerRank Live WebSocket sync, code sandbox OT sync sequence · Docker sandbox security · PostgreSQL + Redis schemas · Replay system · Q&A

Diagram Types (in every design)

Each product README includes multiple mermaid diagrams:

Diagram type Purpose Example
Architecture High-level component layout Client → CDN → LB → Services → DB/Cache/Queue
Sequence Request flow step-by-step Create URL, redirect, book listing, match driver
ER / Data model Entity relationships Users, trips, listings, sessions
Sharding How data is partitioned city_id, user_id, geohash, short_code
State machine Lifecycle transitions Trip: requested → matched → completed
Cache layers CDN → Redis → DB hit rates URL redirect 3-tier cache
Security Isolation, encryption flow Docker sandbox, TLS, PCI

Core Building Blocks (used across all designs)

Concept Used for Details in
Load Balancing Distribute traffic — L4/L7, round robin, sticky sessions (WebSocket) All folders
Sharding Partition DB by user_id, city_id, geohash, short_code Uber, Instagram, URL Shortener
Indexing B-tree, composite, geospatial (PostGIS), inverted (Elasticsearch), GSI (DynamoDB) Uber, Airbnb, URL Shortener
Databases SQL (ACID: trips, bookings) · NoSQL (feeds, URLs, locations) Each folder has full DDL
Caching Redis — feeds, GPS, URL mappings, availability calendars, doc state All folders
API Design REST + GraphQL + WebSocket; pagination, rate limits, idempotency Each folder has API table
Hashing SHA-256 (passwords), Base62 (URL codes), consistent hashing (shards) URL Shortener, Facebook
Encryption TLS in-transit, AES at-rest, PCI for payments Uber, Airbnb
CDN Photos, videos, static assets, hot URL redirects Instagram, Facebook, URL Shortener
Message Queues Kafka — async fan-out, analytics, notifications Uber, Instagram, URL Shortener
CAP Theorem CP for bookings/payments · AP for feeds/likes Airbnb, Instagram

How to approach any interview

flowchart LR
    A[1. Clarify requirements] --> B[2. Estimate capacity]
    B --> C[3. High-level diagram]
    C --> D[4. Data model + APIs]
    D --> E[5. Deep dive bottlenecks]
    E --> F[6. Trade-offs]
Loading 
1. Clarify requirements  →  functional + scale (DAU, QPS, read/write ratio)
2. Estimate capacity     →  storage, bandwidth, servers
3. High-level diagram    →  Client → CDN → LB → Services → DB/Cache/Queue
4. Data model + APIs     →  entities, endpoints, indexing
5. Deep dive             →  bottlenecks, sharding, caching, consistency
6. Trade-offs            →  CAP, SQL vs NoSQL, push vs pull

Scale guide

DAU Stack
1M Monolith + read replicas + Redis
10M Microservices + sharding + CDN + Kafka
100M+ Multi-region + eventual consistency

Quick comparison

Concept Uber Instagram Facebook Airbnb URL Shortener Real-Time Coding
Load Balancer ✅ ✅ ✅ ✅ ✅ ✅ sticky WS
Sharding city_id user_id user_id region hash(code) session_id
Cache Redis GPS Redis feed TAO + Redis Redis cal Redis URLs Redis doc
Queue Kafka Kafka Kafka Kafka Kafka Redis Pub/Sub
SQL PostgreSQL PostgreSQL MySQL PostgreSQL — PostgreSQL
NoSQL Redis Cassandra Cassandra Elasticsearch DynamoDB Redis
Mermaid diagrams 6+ 6+ 6+ 6+ 5+ 5+
Full DB schema ✅ ✅ ✅ ✅ ✅ ✅
Indexing tables ✅ ✅ ✅ ✅ ✅ ✅

Open any folder above for the full detailed design with diagrams, schemas, and interview Q&A.