๐Ÿ“

System Design

High Level Design
Low Level Design
Framework
โšก
HLD

Scalability

Handle more load without breaking
Simple Explanation

Scalability means your system can handle more users or data without crashing or slowing down โ€” like a highway that can add more lanes when traffic grows.

Key Concepts ยท Click to Expand
Vertical Scaling (Scale Up)
โ–ผ
Horizontal Scaling (Scale Out)
โ–ผ
Load Balancer
โ–ผ
Stateless vs Stateful
โ–ผ
๐ŸŒ
Real-World Case Studies
NetflixCASE STUDY ยท 1 / 2

Handling 200M+ subscribers globally

โš  The Problem

When everyone watches Stranger Things at 9 PM on a Friday, traffic spikes 10x. A single server would melt.

โœ“ The Solution

Netflix runs thousands of stateless microservices on AWS that auto-scale horizontally. AWS ELB (Elastic Load Balancer) distributes traffic across servers in multiple availability zones. When CPU usage hits 60%, new instances spin up automatically. When the spike ends, they shut down to save cost.

โš– The Trade-off

Auto-scaling has a 2-5 minute warm-up time. To handle sudden spikes, Netflix keeps a buffer of pre-warmed instances โ€” costing extra but worth it during launches.

โ˜… Interview Takeaway

Stateless services + auto-scaling + load balancers = handle any spike. The key insight: any server can handle any request because no session data is stored locally.

๐Ÿ’ก
Interview Cheat Sheet for Scalability
Always clarify: read-heavy or write-heavy system?
Start with vertical, mention horizontal when limits are hit
Draw the load balancer explicitly in your diagram
Mention auto-scaling triggers (CPU %, request rate, queue depth)