System Design Interview Preparation
“System design interviews test your ability to build real systems, not just solve puzzles.”
🎯 What System Design Interviews Test
System design interviews evaluate your ability to architect large-scale distributed systems. Unlike coding interviews, there’s no single “correct” answer — interviewers want to see your thought process and trade-off analysis.
┌─────────────────────────────────────────────────────────┐
│ SYSTEM DESIGN EVALUATION CRITERIA │
├─────────────────────────────────────────────────────────┤
│ │
│ 1. Problem Exploration (15%) │
│ ├── Requirements gathering │
│ ├── Scope definition │
│ └── Constraints identification │
│ │
│ 2. High-Level Design (25%) │
│ ├── Component identification │
│ ├── Data flow │
│ └── API design │
│ │
│ 3. Deep Dive (35%) │
│ ├── Database schema │
│ ├── Algorithm selection │
│ ├── Scaling strategies │
│ └── Bottleneck resolution │
│ │
│ 4. Trade-offs & Communication (25%) │
│ ├── Pros/cons of decisions │
│ ├── Alternative approaches │
│ └── Clear articulation │
│ │
└─────────────────────────────────────────────────────────┘
📖 In This Section
| Design | Difficulty | Key Concepts |
|---|---|---|
| Design Framework | — | Universal approach to any system design question |
| URL Shortener | ⭐⭐ | Hashing, database, caching, analytics |
| Chat System | ⭐⭐⭐ | WebSockets, message queues, presence, delivery |
| News Feed | ⭐⭐⭐ | Fan-out, ranking, caching, real-time updates |
| Rate Limiter | ⭐⭐ | Algorithms, distributed systems, Redis |
| Key-Value Store | ⭐⭐⭐ | Consistency, replication, partitioning |
| Search Engine | ⭐⭐⭐⭐ | Crawling, indexing, ranking, NLP |
| Video Streaming | ⭐⭐⭐⭐ | CDN, encoding, adaptive bitrate, recommendations |
| Notification System | ⭐⭐⭐ | Multi-channel, delivery guarantees, prioritization |
| Distributed File System | ⭐⭐⭐⭐ | Chunking, replication, consistency, GFS/HDFS |
📊 System Design Concepts Map
┌─────────────────────────────────────────────────────────┐
│ SYSTEM DESIGN CONCEPTS │
├─────────────────────────────────────────────────────────┤
│ │
│ SCALING │
│ ├── Horizontal Scaling (more machines) │
│ ├── Vertical Scaling (bigger machines) │
│ ├── Load Balancing (distribute traffic) │
│ └── Auto-scaling (dynamic capacity) │
│ │
│ DATA │
│ ├── SQL vs NoSQL │
│ ├── Sharding (partition data) │
│ ├── Replication (copy data) │
│ ├── Caching (Redis, Memcached) │
│ └── CDN (static content) │
│ │
│ COMMUNICATION │
│ ├── REST API (synchronous) │
│ ├── Message Queue (async) │
│ ├── WebSockets (real-time) │
│ ├── gRPC (internal services) │
│ └── GraphQL (flexible queries) │
│ │
│ RELIABILITY │
│ ├── Redundancy (no single point of failure) │
│ ├── Failover (automatic recovery) │
│ ├── Circuit Breaker (prevent cascading failures) │
│ └── Retry with Backoff │
│ │
│ CONSISTENCY │
│ ├── Strong Consistency (linearizable) │
│ ├── Eventual Consistency (BASE) │
│ ├── CAP Theorem │
│ └── Consensus (Raft, Paxos) │
│ │
└─────────────────────────────────────────────────────────┘
🎓 When to Use What
Database Selection
SQL (PostgreSQL, MySQL):
├── Structured data with relationships
├── ACID transactions required
├── Complex queries (JOINs)
└── Examples: User accounts, orders, financial data
NoSQL - Document (MongoDB):
├── Flexible schema
├── Nested/hierarchical data
├── Rapid development
└── Examples: Content management, user profiles
NoSQL - Key-Value (Redis, DynamoDB):
├── Simple lookups by key
├── High throughput, low latency
├── Caching, session storage
└── Examples: Cache, session, shopping cart
NoSQL - Wide Column (Cassandra, HBase):
├── Time-series data
├── Write-heavy workloads
├── High availability
└── Examples: Metrics, logs, IoT data
NoSQL - Graph (Neo4j):
├── Relationship-heavy data
├── Social networks
├── Recommendation engines
└── Examples: Friend connections, fraud detection
Communication Pattern Selection
REST API:
├── Client-server communication
├── CRUD operations
├── Stateless, cacheable
└── Use: Public APIs, web apps
gRPC:
├── Internal service communication
├── High performance, streaming
├── Strongly typed (protobuf)
└── Use: Microservice-to-microservice
WebSocket:
├── Real-time bidirectional
├── Persistent connection
├── Low latency
└── Use: Chat, live updates, gaming
Message Queue (Kafka, RabbitMQ):
├── Async processing
├── Decoupling services
├── Buffering load spikes
└── Use: Event processing, task queues
⏱️ Time Management (45-minute interview)
┌─────────────────────────────────────────────┐
│ SYSTEM DESIGN TIME ALLOCATION │
├─────────────────────────────────────────────┤
│ Requirements & Scope 5 min (11%) │
│ High-Level Design 10 min (22%) │
│ Deep Dive Components 20 min (44%) │
│ Trade-offs & Wrap-up 10 min (22%) │
└─────────────────────────────────────────────┘
🔗 Cross-References
- Design Framework — Start here for the universal approach
- Coding Patterns — Algorithm-level patterns
- Architecture Questions — Architecture interview questions
- Cheatsheets — Quick reference for architecture concepts