🎯 System Design Problems → Core Challenge + Key Solution + Trade-offs

---

1. TinyURL

• Core Problem: Unique short URL mapping.
• Key Solution: Base62 encoding + DB storage.

• Trade-offs:

  • Random hash (possible collisions) vs Sequential IDs (predictable, less secure).
  • Store in SQL (consistency) vs NoSQL (scalability).

---

2. Ticket Booking System

• Core Problem: Avoid double booking.
• Key Solution: Pessimistic / Optimistic locking.

• Trade-offs:

  • Pessimistic lock: safe but hurts performance under high load.
  • Optimistic lock: faster but retries needed on conflicts.

---

3. E-commerce (Amazon)

• Core Problem: Product search + inventory consistency.
• Key Solution: Elasticsearch for search + event-driven stock update.

• Trade-offs:

  • Strong consistency (slower checkout) vs Eventual consistency (stock may go negative briefly).
  • Search in DB (simple but slow) vs Search in ES (fast but needs sync).

---

4. OTT Streaming (Netflix)

• Core Problem: Adaptive streaming across bandwidths.
• Key Solution: Transcoding into multiple bitrates + CDN distribution.

• Trade-offs:

  • Pre-transcode all (higher storage) vs Transcode on demand (higher latency).
  • Push video from origin (expensive latency) vs CDN caching (cheap, faster).

---

5. Hotel Booking (Airbnb/MMT)

• Core Problem: Room availability.
• Key Solution: Locks (optimistic/pessimistic) + eventual sync with partners.

• Trade-offs:

  • Strong consistency (slow but safe) vs Eventual consistency (faster, rare overbooking).
  • Central DB (simple, bottleneck) vs Sharded DB (complex, scalable).

---

6. Food Delivery (Zomato/Swiggy)

• Core Problem: Order assignment + live tracking.
• Key Solution: Proximity search (GeoHash, R-Tree) + Pub/Sub updates.

• Trade-offs:

  • Centralized assignment (optimal matching but slow) vs Decentralized (fast but may be sub-optimal).
  • Polling (wasteful, outdated) vs Push via WebSockets (efficient, complex infra).

---

7. Proximity Search

• Core Problem: Fast geo search.
• Key Solution: R-Trees / GeoHash indexing.

• Trade-offs:

  • Exact distance calculation (accurate but slower) vs Approximate GeoHash (fast but may include false positives).

---

8. Chat App (WhatsApp)

• Core Problem: Message delivery guarantees.
• Key Solution: Message queues + ACKs + ordering guarantees.

• Trade-offs:

  • At-least-once (safe but may duplicate) vs At-most-once (no duplicates but may lose).
  • Store messages in DB (durable but slower) vs In-memory queues (fast but risk of loss).

---

9. Distributed Job Scheduler

• Core Problem: Ensure job runs once globally.
• Key Solution: Distributed locks (Redis/Zookeeper) + leader election.

• Trade-offs:

  • Centralized scheduler (simple but SPOF) vs Distributed (resilient but complex).
  • Immediate retries (fast but risk duplicates) vs Delayed retries (safe but slower recovery).

---

10. Collaborative Editor (Google Docs)

• Core Problem: Merge concurrent edits.
• Key Solution: OT (Operational Transformation) or CRDTs.

• Trade-offs:

  • OT (complex but widely adopted) vs CRDT (eventual consistency, simpler conflict handling).
  • Real-time sync (better UX, expensive infra) vs Periodic sync (cheaper, worse UX).

---

11. Ride Sharing (Uber/Ola)

• Core Problem: Matching rider ↔ nearest driver.
• Key Solution: GeoHash indexing + dynamic pricing.

• Trade-offs:

  • Central dispatch (better matches, slower) vs Local matching (faster, less optimal).
  • Strong consistency for driver status (expensive) vs Eventual consistency (rare mismatches).

---

12. Stock Trading (Zerodha/Groww)

• Core Problem: Low-latency order matching.
• Key Solution: In-memory order book + DB persistence.

• Trade-offs:

  • Consistency (safe but slower) vs Availability (fast but may reject valid trades).
  • In-memory matching (fast but volatile) vs DB-based (safe but slower).

---

13. Social Media (Facebook/Instagram)

• Core Problem: News feed generation at scale.
• Key Solution: Fan-out (precompute feeds) vs Fan-in (compute at read time).

• Trade-offs:

  • Fan-out on write (fast reads, heavy writes) vs Fan-out on read (light writes, slow reads).
  • Strong ordering (expensive DB ops) vs Eventual ordering (fast, but minor inconsistencies).

---

✅ Final Tip for Interviews

When asked any design problem, follow this formula:

1. State the core challenge.
2. Propose the main solution.
3. Mention 1–2 trade-offs.

Example (Food Delivery):

“The core challenge is assigning delivery partners efficiently. A common solution is using GeoHash for nearby driver search and Pub/Sub for tracking. The trade-off is centralized assignment gives optimal matches but slower; decentralized assignment is faster but may lead to sub-optimal driver allocation.”

---

---

✅ Quick “Interview Recall Formula”

When asked a system design problem → answer in this format:

• “The core challenge here is X. We solve it using Y.” Example:

“In a ticket booking system, the core challenge is avoiding double booking of seats under concurrency. We solve it using DB locking (pessimistic/optimistic) or distributed locks like Redis with TTL.”

---