WIP

Introduction

Typical Interview Structure

1. Problem scoping: Clarify requirements and constraints of the system
   • Do: Ask clarification questions
   • Don’t: Make assumptions
2. High-level design: Create a rough sketch of the system architecture
   • Do: Identify the main components; draw a diagram
   • Don’t: Get bogged down in details
3. Deep dive: Explore specific components and design choices in detail
   • Do: Explain your rationale; consider bottlenecks
   • Don’t: Afraid to explore unconventional approaches
4. Trade-offs: Discuss the advantages and disadvantages of different approaches
   • Do: Think in terms of scalability, reliability, cost, and complexity
   • Don’t: Dismiss alternatives too quickly
5. Scaling: Address how the system can be scaled to handle increased load
   • Do: Discuss strategies for partitioning data, caching, load balancing, etc
   • Don’t: Forget about real-world constraints

Key System Design Concepts

Core Design Principles

• SPARCS
  • Scalability
  • Performance
  • Availability
  • Reliability
  • Consistency
  • Security
• Scalability: Ability of a system to handle a growing amount of work

  • 2 types:

    • Horizontal scalability: Adding more machines to the system to distribute the workload
      • Good for flexibility and cost-effectiveness
    • Vertical scalability: Upgrading single machine’s resources, such as adding RAM or CPU cores
      • Easier to implement, but can be limited by machine’s maximum capacity

  • Scalability techniques:

    • Sharding: Distributes data across multiple servers in smaller units

      

    • Replication: Creates and maintains copies of data across multiple servers

      

    • Partitioning: Divides a database into smaller, organized segments

      

    • Caching: Improves response times through efficient retrieval of frequently accessed data

      

  • How to achieve it:

    • Modular design
    • Load balancing
    • Caching
    • CDNs
    • Elasticity
    • Asynchronous processing

  • What to avoid:

    • Monolithic architectures
    • Stateful components
    • No load testing

  • Key principles to consider

    • CAP theorem
    • Microservices
    • Event-driven architecture
    • ACID transaction
• Performance: How quickly and efficiently your system can get things done
  • 3 types:
    • Caching: Data in memory means you don’t have to dig for it in slower storage
    • Load balancing: Distribute the workload evenly
    • Optimizing algorithms: Choosing right alg or data structure can amke the difference
• Availability: Measures how often your system is up and running

  • 3 types:

    • Monitoring
    • Load balancing
    • Failover mechanisms

  • Nines of availability

    Availability %	Downtime per year	Downtime per month
    90% (1 nine)	36.5 days	72 hours
    99% (2 nines)	3.65 days	7.20 hours
    99.9% (3 nines)	8.76 hours	43.8 minutes
    99.99% (4 nines)	52.56 minutes	4.32 minutes
    99.999% (5 nines)	5.26 minutes	25.9 seconds

  • Prioritizing availability is win-win → good for users and system

• Reliability: Ability of system to function correctly, even in face of failures
  • 2 types:
    • Fault tolerance: Designing systems that can continue even when some components fail
      • Involves redundancy, where multiple components perform the same function so that others can take over if one fails
    • Redundancy: Duplicating critical components or data to ensure the system can still function if one fails
      • Can be done at various levels
• Consistency: Ensures all parts of your system agree on the data state; tricky in distributed systems, where data is spread across multiple machines
• Security

Building Blocks of System Design