System Design Interview Cheat Sheet — Save This Page
The one reference page you need before walking into any system design interview. Capacity estimation formulas, database selection guide, caching patterns, CAP theorem, and 30+ design decisions distilled to their essentials.
After reading: You'll have every key formula, number, and trade-off pattern on a single reference you can revisit the night before any interview.
Bookmark this page. Come back to it before every mock interview and every real interview. This is the consolidated reference for everything you need to carry into a system design interview.
Capacity Estimation Formulas
The core formula:
Peak QPS = (Daily Requests ÷ 86,400) × Peak Multiplier
Storage/year = Writes/day × Object size (bytes) × 365
Bandwidth = Peak QPS × Avg object size (bytes)
Cache size = Read QPS × Avg response × Cache hit ratio × TTL
Common peak multipliers:
- Normal product: 2–3×
- Flash sale / live match: 10–20×
- Morning rush (food delivery): 5×
Object size reference:
Short text (tweet, SMS): ~200 bytes
User record: ~500 bytes
Product listing (e-commerce): ~1–2 KB
Small image (thumbnail): ~50 KB
Medium photo: ~500 KB
1 min compressed video: ~50 MB
Worked estimates for common problems:
| Problem | Daily Writes | Peak QPS | Storage/yr | |---|---|---|---| | Twitter (500M tweets/day) | 500M | ~17K (peak 50K) | ~50TB | | WhatsApp (100B msgs/day) | 100B | ~1.2M | ~10PB | | Swiggy (5M orders/day) | 5M | ~170 (peak 1K) | ~10GB | | YouTube (100K uploads/day, 10GB avg) | 100K | ~1 | ~350PB/yr |