System Design Expert¶

The system design framework¶

The system design framework — a key concept in System Design.

1. REQUIREMENTS   — What exactly are we building? (functional + non-functional)
2. ESTIMATION     — How much data? How many users? QPS?
3. HIGH-LEVEL     — Major components and how they connect
4. DETAILED       — Dive into the hardest parts
5. TRADE-OFFS     — Why this design over alternatives?

Example: Design a URL shortener¶

Example: Design a URL shortener in System Design — what it is and when to use it.

Requirements¶

• Shorten a URL → return short code (7 chars)
• Redirect short code → original URL
• 100M URLs, 1000 reads/sec, 10 writes/sec
• Analytics (click count)

High-level design¶

Client → Load Balancer → API Server → Database
                              ↓
                         Cache (Redis)

Data model¶

# URL mapping
class URL:
    short_code: str    # "abc1234" (primary key)
    original_url: str  # "https://very-long-url.com/..."
    created_at: datetime
    click_count: int
    user_id: int | None

Core algorithm¶

import hashlib
import string

ALPHABET = string.ascii_letters + string.digits   # 62 chars
BASE = len(ALPHABET)

def encode_id(num: int) -> str:
    """Convert numeric ID to base-62 string."""
    if num == 0:
        return ALPHABET[0]
    result = []
    while num:
        result.append(ALPHABET[num % BASE])
        num //= BASE
    return "".join(reversed(result))

def shorten(original_url: str) -> str:
    # Option 1: Auto-increment ID → base62
    url_id = db.insert(original_url)   # returns auto-inc ID
    return encode_id(url_id)

    # Option 2: Hash-based
    hash_hex = hashlib.md5(original_url.encode()).hexdigest()
    return hash_hex[:7]   # collision possible!

Scaling considerations¶

Key trade-offs to discuss¶

Key trade-offs to discuss in System Design — what it is and when to use it.

Common system design problems¶

Common system design problems in System Design — what it is and when to use it.

Practice Exercises¶

1. Design a Twitter-like feed — handle writes from 100K users, reads from 10M users.
2. Design a chat application — 1:1 and group chat with read receipts and typing indicators.
3. Design a job queue — reliable, distributed, with retry and dead-letter queue.
4. Design a rate limiter — multiple algorithms (fixed window, sliding window, token bucket).
5. Design a notification service — email, push, SMS with batching and user preferences.