Case Study: URL Shortener

Let's design a URL shortener like Bitly or TinyURL. This is a classic system design interview question that touches on every concept we've learned.

We'll follow the Architect's Framework: Clarify → Estimate → Design → Deep-Dive.

Step 1: Clarify Requirements

Functional Requirements

What does the system need to do?

Non-Functional Requirements

How well does the system need to perform?

Step 2: Back-of-Envelope Estimation

Traffic

URLs created per month: 100 million
URLs created per second: 100M / (30 * 24 * 3600) ≈ 40 writes/sec

Read:Write ratio: 100:1
Redirects per second: 40 * 100 = 4,000 reads/sec

Storage

URL length (average): 500 bytes
Short code: 7 bytes
Metadata: ~100 bytes
Total per URL: ~600 bytes

URLs stored for 5 years: 100M * 12 * 5 = 6 billion URLs
Storage needed: 6B * 600 bytes = 3.6 TB

Summary

Step 3: High-Level Design

API Design

POST /api/shorten
Body: { "url": "https://example.com/very/long/url" }
Response: { "short_url": "https://short.ly/abc1234" }

GET /{short_code}
Response: 302 Redirect to original URL

Step 4: Deep Dive

Short Code Generation

How do we generate unique 7-character codes?

#### Option 1: Hash the URL

short_code = base62(md5(long_url))[:7]

#### Option 2: Counter + Base62

# Global counter: 1, 2, 3, 4...
short_code = base62(counter)

#### Option 3: Pre-generated Keys (Recommended)

Generate keys offline, store in a key service. Each API server grabs a batch.

Base62 Encoding

Why 7 characters in Base62?

Base62 characters: a-z (26) + A-Z (26) + 0-9 (10) = 62

7 characters: 62^7 = 3,521,614,606,208 (3.5 trillion)

At 100M URLs/month = 1.2B/year
Time to exhaust: 3.5T / 1.2B = 2,917 years

Database Schema

CREATE TABLE urls (
    short_code VARCHAR(7) PRIMARY KEY,
    long_url TEXT NOT NULL,
    user_id UUID,
    created_at TIMESTAMP DEFAULT NOW(),
    expires_at TIMESTAMP,
    click_count INT DEFAULT 0
);

-- Index for expiration cleanup
CREATE INDEX idx_expires ON urls(expires_at);

Caching Strategy

def redirect(short_code):
    # 1. Check cache
    long_url = cache.get(short_code)
    if long_url:
        return redirect_302(long_url)  # Cache hit!

    # 2. Cache miss - check database
    long_url = db.get(short_code)
    if not long_url:
        return 404

    # 3. Update cache
    cache.set(short_code, long_url, ttl=3600)

    return redirect_302(long_url)

With 80% hit rate:

• 4,000 reads/sec × 0.2 (misses) = 800 DB queries/sec

Analytics (Async)

Don't block redirects to record analytics. Use a message queue:

def redirect(short_code):
    long_url = get_url(short_code)  # From cache or DB

    # Async analytics
    queue.publish({
        'short_code': short_code,
        'timestamp': now(),
        'referrer': request.headers.get('Referer'),
        'user_agent': request.headers.get('User-Agent'),
        'ip': request.ip
    })

    return redirect_302(long_url)

Workers consume from the queue and batch-write to an analytics database.

Complete Architecture

Design Decisions Summary

Key Takeaways

Next up: HLD Assessment - Test your knowledge across all topics.