System Design Tutorial Notes

Overview

• Key topics: scalability, reliability, data handling, high-level architecture
• Focus: system design interviews, core concepts, practical strategies
• Importance: not about coding but about integrating an entire system

Computer Architecture Basics

• Binary & Data Units: Bits (0 or 1), Bytes (8 bits), KB, MB, GB, TB
• Disk Storage: HDD vs SSD, non-volatile, stores OS, applications, user files
  • HDD: 80-160 MBps
  • SSD: 500-3500 MBps
• RAM: Volatile, stores active data, faster than SSD
  • Speed: >5000 MBps
  • Size: GB in consumer devices, hundreds of GB in servers
• Cache: Smaller, faster access than RAM, used for frequently accessed data
  • L1, L2, L3 cache levels
  • Access time in nanoseconds
• CPU: Processes code, fetches/decodes/executes instructions, interacts with RAM, disk, cache
• Motherboard: Connects all components, allows data flow

High-Level Architecture of Production-Ready App

• CI/CD Pipeline: Automates deployment, tools like Jenkins, GitHub Actions
• Load Balancers & Reverse Proxies: Manage user requests, ensure even distribution, e.g., NGINX
• External Storage Server: Keeps data separate from production server
• Logging & Monitoring: Tools like PM2 (backend), Sentry (frontend)
• Alerting Services: Detect anomalies, send notifications, e.g., via Slack
• Debugging: Use logs, replicate issue in test environment, hotfixes

System Design Principles

• Good Design Principles: Scalability, Maintainability, Efficiency
• Robust & Resilient: Plan for failures, ensure performance even during issues
• Key Elements: Moving data, storing data, transforming data

CAP Theorem (Brewer's Theorem)

• Consistency: All nodes have same data
• Availability: System is always operational
• Partition Tolerance: System functions despite network partitions
• Trade-offs: Only 2 of 3 properties can be achieved simultaneously

Measuring System Performance

• Availability: Measured in uptime/downtime, aim for 99.999%
• Reliability & Fault Tolerance: Ensure system works correctly and handles failures
• Throughput & Latency: Measures data handling over time vs. time to complete single request

Networking Basics

• IP Addressing: IPv4 (32-bit), IPv6 (128-bit)
• Data Packets: Include IP headers, application layer data
• Transport Layer Protocols: TCP (reliable), UDP (faster but less reliable)
• DNS (Domain Name System): Translates domain names to IP addresses
• Networking Infrastructure: Public vs. private IPs, static vs. dynamic IPs
• Ports: Identify specific processes/services

Application Layer Protocols

• HTTP: Stateless, request-response, status codes (200, 300, 400, 500)
• WebSockets: Two-way communication
• SMTP, IMAP, POP3: Email transmission and retrieval
• FTP, SSH: File transfer, secure remote operations
• WebRTC, MQTT, AMQP: Real-time communication, messaging
• RPC: Remote procedure calls for code execution on remote machines

API Design

• CRUD Operations: Create (POST), Read (GET), Update (PUT/PATCH), Delete (DELETE)
• Protocols: HTTP, WebSockets, JSON, XML, Protocol Buffers
• APIs Paradigms: REST, GraphQL, gRPC
• Versioning, Rate Limiting, CORS: Maintain backward compatibility, prevent misuse

Caching & CDNs

• Browser Caching: Stores website resources locally
• Server Caching: Reduces database queries
• Database Caching: Improves read-heavy applications
• CDNs: Serve static content from geographically distributed servers

Proxy Servers

• Types: Forward, Reverse, Open, Transparent, Anonymous, Distorting, High Anonymity
• Use Cases: Caching, load balancing, security, anonymizing
• Load Balancing Strategies: Round-robin, least connections, weighted, geographical, consistent hashing

Database Essentials

• Types: Relational (SQL), NoSQL, In-Memory
• Scaling: Vertical (scale-up), Horizontal (sharding, replication)
• Performance Techniques: Caching, indexing, query optimization
• CAP Theorem: Prioritize two out of consistency, availability, partition tolerance