Review system design fundamentals, including abstraction and non-functional characteristics in designing modern systems.

Deep Dive into System Design Interview

System design is the process of defining components and their integration, APIs, and data models to build large-scale systems. Mobile and cloud computing, combined with expanded internet access, make system design a core skill for the modern developer. This module provides a brief overview of the concepts required to design scalable systems.

System Design Essentials

Estimating the Number of Servers



- Total requests per second $\frac{500M \times 20}{86400} = 115K$ requests per second.
- For the number of servers: $\frac{Number\ of\ Requests/sec}{RPS\ of\ server} = \frac{115K}{64000} \approx 2$ servers to handle all the requests.

When we divided the daily user requests by the number of seconds in a day, we assumed that those requests were uniformly distributed over those seconds.

This raises the question of whether the above assumption is reasonable. Rarely will a service get such a nice, uniform distribution of requests.

Doing the above is a simple strategy; what we get is a lower bound on the respective resource.

When we divided the daily user requests by the number of seconds in a day, we assumed that those requests were uniformly distributed over those seconds.
This raises the question of whether the above assumption is reasonable. Rarely will a service get such a nice, uniform distribution of requests.
Doing the above is a simple strategy; what we get is a lower bound on the respective resource.

Can you identify a hidden assumption in our calculations above?

Can you identify a hidden assumption in our calculations above?

Points to ponder

Points to ponder

\text{Servers needed at peak load} = \frac{\text{Number of requests/second}}{\text{RPS of server}} = \frac{10 \ \text{Billion}}{64,000} = 157\text{K}


Explore practical resource estimation by applying BOTECs to a high-scale service. Quantify server count, storage needs, and bandwidth requirements based on traffic assumptions. Learn to adjust peak load estimates using principles such as the Pareto principle to ensure System Design feasibility and manage costs.

Daily active users (DAU)	500	Million
Requests on average / user / day	20
Total requests / day	f10	Billion
Total requests / second	f115	K
Total servers required	f2

Examples of Resource Estimation

Introduction

Number of servers required

Estimating the Number of Servers

Peak capacity

Improving the RPS of a server