System requirements

Functional:

The system should:

1. Allow users to create a short URL from a long URL
2. Redirect a short URL to the original long URL
3. Generate unique short codes for each long URL
4. Support custom aliases
5. Prevent duplicate short codes
6. Handle invalid / expired URLs gracefully
7. Support URL expiration (optional)
8. Track basic analytics (click count, timestamp – optional)

Non-Functional:

The system should be:

1. Highly available (URL redirection must not fail)
2. Low latency (redirect in < 50 ms)
3. Scalable (handle millions of URLs & redirects)
4. Consistent (same short URL always maps to same long URL)
5. Durable (URL mappings should not be lost)
6. Secure (prevent malicious URLs, rate limiting)
7. Fault tolerant (cache / DB failures handled gracefully)

API design

POST /api/v1/shorten

{

"longUrl": "https://example.com/very/long/url",

"customAlias": "myurl", // optional

"expiry": "2026-01-01" // optional

}

{

"shortUrl": "https://tiny.ly/myurl"

}

GET /{shortCode}

GET /api/v1/analytics/{shortCode}

High-level design

Core Components

1. Client
   • Browser / Mobile App
2. Load Balancer
   • Distributes traffic across servers
3. API Server
   • Handles shorten & redirect logic
   • Stateless
4. Cache (Redis)
   • Stores shortCode → longURL
   • Reduces DB load
5. Database
   • Persistent storage of mappings
6. ID Generator
   • Generates unique short codes (Base62)

Detailed component design

1️⃣ Short Code Generator

Approach: Base62 Encoding

• Characters: a-z A-Z 0-9 → 62 chars
• Generate incremental ID:

Why Base62?

• Short URLs
• URL safe
• Huge combinations

Scalability

• Can generate billions of unique URLs
• ID generation can be centralized or Snowflake-style

2️⃣ Cache (Redis)

Why Cache?

• 90%+ traffic is read-heavy
• Redirection must be fast

Data Stored

Strategy

• Read-through cache
• TTL = URL expiry
• Cache miss → DB → populate cache

3️⃣ API Server

Responsibilities

• Validate URLs
• Generate short code
• Store mapping
• Handle redirects

Scaling

• Stateless → horizontal scaling
• Auto-scale behind load balancer

Database design