Scalability — Overview#
Scalability is not one decision — it's a sequence of bottleneck, fix, next bottleneck.
A system that works for 100 users will break for 1 million. This folder covers how to identify where it breaks, how vertical and horizontal scaling differ, and the mechanisms that make horizontal scaling work — load balancing and auto-scaling. Database scaling and caching are deep topics covered in their own dedicated phases.
Files in this folder#
| File | Topic |
|---|---|
| 01-Scalability.md | What scalability is, vertical vs horizontal, the three bottlenecks |
| 02-Load-Balancing/ | How traffic is distributed across servers — algorithms, L4, L7, API Gateway |
| 03-Auto-Scaling/ | Feedback loop, reactive vs predictive, connection draining, cold start |
| 04-Interview-Cheatsheet.md | How to answer the "now scale it 10x" question |