System Design Examples
Step-by-step walkthroughs of 27 system design interview questions — from classic URL shortener to distributed file systems.
Recommended Study Order
Tip
Don't study randomly. Follow this progression — each design builds on concepts from earlier ones.
Tier 1: Foundation (Start Here)
| # |
Design |
Why First |
Prerequisites |
| 1 |
URL Shortener |
Simplest end-to-end design |
Databases, Caching |
| 2 |
Rate Limiter |
Core API protection pattern |
Redis, distributed systems |
| 3 |
Distributed Cache |
Appears in every other design |
Consistent hashing |
| 4 |
Key-Value Store |
Deepens distributed systems understanding |
Replication, consensus |
Tier 2: Communication & Social
Tier 3: Media & Content
Tier 4: Real-time & Geospatial
Tier 5: Infrastructure & Commerce
| # |
Design |
Builds On |
New Concepts |
| 15 |
Task Scheduler |
Distributed Cache, KV Store |
Lease-based execution, priority queues |
| 16 |
Distributed Message Queue |
Task Scheduler, KV Store |
Append-only log, consumer groups, zero-copy I/O |
| 17 |
Event Booking (Ticketmaster) |
Rate Limiter |
Inventory locking, flash sales |
| 18 |
Payment System |
Event Booking |
Double-entry ledger, idempotency |
| 19 |
Metrics & Monitoring |
Message Queue |
Time-series storage, alerting, Gorilla compression |
| 20 |
Email Delivery System |
Notification System |
SMTP, DKIM/SPF, IP reputation, deliverability |
Tier 6: Data Infrastructure
| # |
Design |
Builds On |
New Concepts |
| 21 |
Distributed File System (GFS) |
KV Store, Cache |
Master-chunk, leases, replication, append-only writes |
| 22 |
Ad Click Aggregator |
Message Queue, Metrics |
Real-time aggregation, exactly-once, click fraud, reconciliation |
Tier 7: Search
| # |
Design |
Builds On |
New Concepts |
| 23 |
Web Crawler |
Message Queue, Cache |
URL frontier, politeness, dedup |
| 24 |
Search Autocomplete |
Cache, Web Crawler |
Trie, ranking, type-ahead |
Tier 8: Modern & Trending
| # |
Design |
Builds On |
New Concepts |
| 25 |
Gaming Leaderboard |
Cache, Message Queue |
Redis sorted sets, real-time ranking, anti-cheat |
| 26 |
API Gateway |
Rate Limiter, Load Balancer |
Plugin architecture, circuit breaker, hot config reload |
| 27 |
Content Delivery Network |
Distributed Cache, Cloud Storage |
Edge caching, PoP hierarchy, Anycast routing |
Staff Engineer (L6) Track
Preparing for a Staff / Principal / L6 role? Start here.
Important
At L6, the interviewer gives you a vague prompt and expects you to define the problem, drive the whiteboard, and discuss multi-year evolution. The designs below include dedicated "Staff Engineer Deep Dive" sections.
Must-Read First
Priority Design Problems (80/20 Rule)
These 5 designs cover 80% of distributed systems concepts tested at L6:
| Design |
Staff-Level Concepts Covered |
| Key-Value Store |
CAP theorem, consistent hashing, quorum, vector clocks, Spanner/TrueTime, multi-region replication |
| Rate Limiter |
Global rate limiting, race conditions, cascading failures, load shedding, adaptive limits |
| Collaborative Editor |
OT vs CRDTs decision framework, WebSocket scaling, hot document problem, multi-region |
| Task Scheduler |
Fencing tokens, zombie workers, multi-tenant fairness, cron correctness at scale |
| Notification System |
Exactly-once delivery chain, transactional outbox, load shedding |
Supporting Advanced Topics
How to Use These Examples
Each example follows a consistent structure that mirrors what interviewers expect:
| Phase |
What You'll Learn |
Time in Interview |
| 1. Requirements |
Clarifying questions to ask |
5 minutes |
| 2. Estimation |
Back-of-envelope calculations |
5 minutes |
| 3. High-Level Design |
Architecture overview |
10 minutes |
| 4. Deep Dive |
Key components in detail |
15 minutes |
| 5. Scaling & Trade-offs |
Production considerations |
5 minutes |
Tip
Practice drawing these designs on a whiteboard or paper. The physical act of drawing helps with memory and interview confidence.
All Designs by Category
Infrastructure & Data
| Design |
Difficulty |
Core Pattern |
| URL Shortener |
⭐⭐ Medium |
Hashing, distributed IDs |
| Rate Limiter |
⭐⭐ Medium |
Token bucket, sliding window |
| Key-Value Store |
⭐⭐⭐⭐ Hard |
Consistent hashing, quorum |
| Distributed Cache |
⭐⭐⭐ Medium-Hard |
LRU, hot keys, stampede |
| Distributed Message Queue |
⭐⭐⭐⭐ Hard |
Append-only log, consumer groups |
| Task Scheduler |
⭐⭐⭐ Medium-Hard |
Priority queue, leases |
| Metrics & Monitoring |
⭐⭐⭐⭐ Hard |
Time-series, alerting |
| Distributed File System (GFS) |
⭐⭐⭐⭐ Hard |
Master-chunk, leases, replication |
| Ad Click Aggregator |
⭐⭐⭐⭐ Hard |
Real-time aggregation, exactly-once |
Communication & Social
Media & Content
Real-time & Geospatial
Commerce & Finance
Search
Modern & Trending
Pattern Recognition
| Pattern |
Where You'll See It |
| Cache-aside |
URL Shortener, Rate Limiter, Distributed Cache, Proximity Service |
| Message Queue |
Voting System, Notification, Video Streaming, Task Scheduler, Message Queue |
| Read Replicas |
URL Shortener, Voting System, News Feed |
| Distributed IDs |
URL Shortener, Payment System — Snowflake algorithm |
| Idempotency |
Voting, Notification, Payment, Task Scheduler, Message Queue |
| Rate Limiting |
Rate Limiter, Web Crawler, Event Booking (virtual queue) |
| Consistent Hashing |
Key-Value Store, Distributed Cache — data partitioning |
| WebSockets |
Chat System, Collaborative Editor, Ride Sharing |
| Geospatial Indexing |
Ride Sharing, Proximity Service — geohash, quadtree |
| Fan-out |
News Feed, Photo Sharing — push vs pull vs hybrid |
| State Machine |
Payment System, Ride Sharing, Task Scheduler, Event Booking |
| CDN |
Video Streaming, Photo Sharing, Cloud Storage |
| Conflict Resolution |
Key-Value Store (vector clocks), Collaborative Editor (OT/CRDT) |
| Append-Only Log |
Message Queue, Event Sourcing, Metrics & Monitoring, Distributed File System |
| Stream Processing |
Ad Click Aggregator, Metrics & Monitoring — windowed aggregation |
| Master-Worker |
Distributed File System, Task Scheduler — coordination patterns |
| Sorted Sets / Skip Lists |
Gaming Leaderboard — real-time rank queries |
| Circuit Breaker |
API Gateway, Notification System — fault isolation |
| Edge Caching |
Content Delivery Network, Video Streaming — PoP hierarchy |
| Plugin / Middleware |
API Gateway — extensible request processing |
Note
Master these patterns and you can apply them to any new problem the interviewer throws at you.
Quick Reference: Complexity
| Design |
Read/Write Ratio |
Scale Challenge |
Core Trade-off |
| URL Shortener |
100:1 (read-heavy) |
Billions of URLs |
Consistency vs latency |
| Rate Limiter |
N/A |
Millions of clients |
Precision vs memory |
| Key-Value Store |
Varies |
Partitioning |
Consistency vs availability |
| Distributed Cache |
Read-heavy |
Hot keys |
Stale reads vs latency |
| Message Queue |
Write-heavy |
1M+ msgs/sec |
Ordering vs throughput |
| Metrics & Monitoring |
Write-heavy |
100K metrics/sec |
Granularity vs storage |
| Notification System |
Write-heavy |
Millions/minute |
Reliability vs latency |
| Web Crawler |
N/A |
Billions of pages |
Speed vs politeness |
| Chat System |
Write-heavy |
Millions sockets |
Ordering vs fan-out |
| News Feed |
Read-heavy |
Billions loads/day |
Fan-out cost vs read latency |
| Search Autocomplete |
Read-heavy |
Peak QPS |
Freshness vs latency |
| Voting System |
1:1 |
Spike traffic |
Accuracy vs throughput |
| Video Streaming |
Read-heavy |
Petabytes of video |
Storage cost vs quality |
| Photo Sharing |
Read-heavy |
Billions of images |
Fan-out cost vs latency |
| Collaborative Editor |
Write-heavy |
Concurrent editors |
Consistency vs responsiveness |
| Ride Sharing |
Write-heavy |
Millions of drivers |
Speed vs optimality |
| Cloud Storage |
Balanced |
Petabytes |
Sync speed vs bandwidth |
| Event Booking |
Write-heavy (flash) |
Spike traffic |
Consistency vs throughput |
| Task Scheduler |
Write-heavy |
Millions of tasks |
At-least-once vs exactly-once |
| Payment System |
Write-heavy |
Financial accuracy |
Consistency vs availability |
| Proximity Service |
Read-heavy |
Millions locations |
Precision vs query speed |
| Distributed File System |
Write-heavy (append) |
Petabytes |
Single master vs availability |
| Ad Click Aggregator |
Write-heavy |
1M events/sec |
Exactness vs latency |
| Gaming Leaderboard |
Read-heavy |
50K writes/sec, 500K reads/sec |
Rank precision vs latency |
| API Gateway |
Balanced |
100K RPS/node |
Routing overhead vs feature richness |
| Content Delivery Network |
Read-heavy |
10M RPS globally |
Cache hit ratio vs freshness |
What's Next?
After mastering these software system designs:
- Go deeper with Advanced Topics for Senior/Staff-level concepts
- Learn ML fundamentals in GenAI/ML Fundamentals — 7 building blocks
- Tackle ML designs in ML System Design — 10 production ML systems
- Master GenAI in GenAI System Design — 10 LLM/GenAI systems with interview transcripts