System Design Basics - Interview
Guide
Scalability, load balancing, caching, databases, queues and a reusable design
framework
Complete Study Guide | Key Concepts | Examples | Practice Questions
, 1. The System Design Interview Framework
Interviewers score structured thinking, not perfect answers. Use this framework for any 'design X' question.
- 1. Clarify: functional + non-functional requirements, scale (users, reads/writes, latency).
- 2. Estimate: back-of-envelope math (QPS, storage, bandwidth) - shows engineering judgement.
- 3. High-level design: client -> CDN -> load balancer -> app servers -> DB/cache.
- 4. Deep dive: pick 2-3 hard parts (e.g. feed ranking, id generation) and go deep.
- 5. Bottlenecks: name what breaks at 10x and how you'd fix it.
- Always state assumptions out loud; there is rarely one right answer.
TIP: Numbers to memorise: 1M QPS is huge; 1KB x 1M users/day ~ 1GB/day; SSD ~ 0.1ms, network ~
1-100ms.
Page 1
Guide
Scalability, load balancing, caching, databases, queues and a reusable design
framework
Complete Study Guide | Key Concepts | Examples | Practice Questions
, 1. The System Design Interview Framework
Interviewers score structured thinking, not perfect answers. Use this framework for any 'design X' question.
- 1. Clarify: functional + non-functional requirements, scale (users, reads/writes, latency).
- 2. Estimate: back-of-envelope math (QPS, storage, bandwidth) - shows engineering judgement.
- 3. High-level design: client -> CDN -> load balancer -> app servers -> DB/cache.
- 4. Deep dive: pick 2-3 hard parts (e.g. feed ranking, id generation) and go deep.
- 5. Bottlenecks: name what breaks at 10x and how you'd fix it.
- Always state assumptions out loud; there is rarely one right answer.
TIP: Numbers to memorise: 1M QPS is huge; 1KB x 1M users/day ~ 1GB/day; SSD ~ 0.1ms, network ~
1-100ms.
Page 1
