System Design: FAANG Interview Ready

DNS resolution, TCP vs UDP, HTTP/1.1 vs HTTP/2 vs HTTP/3, TLS, WebSockets, and long polling. The building blocks every system design answer depends on.

• Cloudflare Learning Center: How the Internet Worksarticlefree
• ByteByteGo: EP27 - How does HTTPS work?articlefree
• High Scalability: All Time Favoritesarticlefree

• Explain DNS lookup with all hops
• TCP vs UDP: when to use which
• HTTP/2 multiplexing vs HTTP/1.1 pipelining
• WebSocket handshake and use cases

ACID, BASE, CAP theorem, consistency models, indexing, sharding, replication, and when to choose PostgreSQL vs Cassandra vs DynamoDB vs Redis.

• CMU Database Course: Andy Pavlo (free lectures)coursefree
• Martin Kleppmann: Designing Data-Intensive Applications (DDIA)coursepaid
• ByteByteGo: A Crash Course in Databasesarticlefree

• CAP theorem: pick 2 of 3 - worked example
• B-tree vs LSM-tree index tradeoffs
• Read replica lag and consistency windows
• Sharding strategy: hash vs range vs directory

CPU caches, application-level caching, Redis strategies (cache-aside, write-through, write-behind), CDN architecture, and cache invalidation.

• ByteByteGo: A Crash Course in Cachingarticlefree
• Redis Best Practicesdocfree
• AWS: Caching Best Practicesdocfree

• Cache-aside pattern with TTL
• Cache stampede: prevention strategies
• CDN: push vs pull, origin shield
• Design a caching layer for a Twitter timeline

L4 vs L7 load balancers, algorithms (round-robin, least-connections, consistent hashing), reverse proxies, service meshes, and API gateways.

• NGINX: Load Balancing Guidedocfree
• ByteByteGo: Consistent Hashing Explainedarticlefree

• Consistent hashing ring - code it
• Health check and circuit breaker pattern
• Sticky sessions vs stateless design

Kafka, RabbitMQ, SQS: pub/sub, fan-out, at-least-once vs exactly-once delivery, event sourcing, and CQRS.

• Confluent: Event Streaming Patternsdocfree
• DDIA Chapter 11: Stream Processing (book)coursepaid
• ByteByteGo: Kafka vs RabbitMQ vs SQSarticlefree

• At-least-once delivery: idempotency design
• Dead letter queue and retry strategy
• CQRS: separate read and write models

Two-phase commit, Paxos, Raft consensus, distributed locks with Redis/Zookeeper, leader election, and saga pattern for distributed transactions.

• The Raft Consensus Algorithm (visual)linkfree
• Martin Fowler: Patterns of Distributed Systemsarticlefree
• DDIA Chapter 9: Consistency and Consensuscoursepaid

• Raft leader election - explain each phase
• Distributed lock with Redis SETNX + expiry
• Saga pattern: choreography vs orchestration

Service decomposition, API gateway, service discovery, circuit breaker, bulkhead, and gRPC vs REST vs GraphQL for inter-service communication.

• microservices.io: Patterns by Chris Richardsonarticlefree
• Google API Design Guidedocfree
• ByteByteGo: REST vs GraphQL vs gRPCarticlefree

• Decompose a monolith by bounded context
• Circuit breaker: closed, open, half-open states
• Design an API gateway with rate limiting

URL shortener, rate limiter, web crawler, notification system, news feed, search autocomplete. Practice the ByteByteGo framework: requirements, estimation, high-level, deep dive.

• Alex Xu: System Design Interview Vol 1 (book)coursepaid
• Alex Xu: System Design Interview Vol 2 (book)coursepaid
• ByteByteGo: System Design Newsletter + Videoscoursepaid
• System Design Primer (GitHub, free)repofree
• Exponent: System Design Interview Practicecoursepaid

• Design TinyURL (URL shortener)
• Design a distributed rate limiter
• Design Twitter home timeline (fan-out on write vs read)
• Design Google Drive / Dropbox
• Design a notification system (push, email, SMS)
• Design YouTube / Netflix CDN architecture

Live leaderboards, collaborative editing, ride-sharing dispatch, payments, and recommendation engines. Covers Operational Transformation, CRDTs, and stream processing at scale.

• Alex Xu: System Design Interview Vol 2 (advanced)coursepaid
• Engineering at Meta: Distributed Systems (blog)articlefree
• Google Engineering Blogarticlefree

• Design a real-time collaborative editor (Figma/Google Docs)
• Design Uber's dispatch system
• Design a live sports leaderboard
• Design a payment system with idempotency

Timed 45-minute mock interviews with peer or AI feedback. Practice pacing, estimation, and whiteboarding communication.

• interviewing.io: System Design Mock Interviewslinkfree
• Pramp: Free Peer System Design Interviewslinkfree
• IGotAnOffer: System Design Frameworkarticlefree

• 5 timed 45-min mock sessions completed
• Receive and action feedback on communication gaps
• Whiteboard a system end-to-end without notes