System requirements

Functional:

Given a URL, we'll compress it into a shorter one

Non-Functional:

Loadbalancer from clientside to server

Shard database by first character by URL's root domain

Use replicas -> If url has already been shortened then grab existing shortened URL

Capacity estimation

URL length 2000-3000 characters long

API design

External:

• /shortenURL
• Take in original URL as a parameter
• Read from replica to see if URL exists, if not add into DB

Database design

Shorted URL Table:

ShortenedURL(Str) -> Hashed from backend

OriginalURL(Str, PK) -> Passed to API from clientside

High-level design

User inputs a URL, clientside sends to server

Serverside hashes the URL and stores in DB

Request flows

User -> Clientside -> Server

Detailed component design

Server side:

• Read from DB
• If DB has URL: grab shortened URL
• if DB does not have URL: Hash URL and add into DB

Trade offs/Tech choices

Explain any trade offs you have made and why you made certain tech choices...

Failure scenarios/bottlenecks

Too many requests at time or might be straining on DB to have so many transactions with it

This is why we use a replica

Future improvements

What are some future improvements you would make? How would you mitigate the failure scenario(s) you described above?