Vertical vs. Horizontal Scaling & Stateful vs. Stateless Servers
Experiment with incoming user traffic below. Watch what happens when you scale Vertically (Scale Up) on a single monster box versus Horizontally (Scale Out) across multiple commodity nodes—and toggle Session State to see why sticky local sessions break horizontal elasticity.
1. Architecture Controls
⚠️ Notice: Even Tier 5 hits a physical motherboard ceiling at 12,000 RPS and requires a reboot window to upgrade!
Sessions locked inside individual server memory.
Initializing diagnostic telemetry...
Curriculum Progress & Top Free Study Resources
🔥 Core YouTube Channels & Open-Source Guides (Direct Links)
Gaurav Sen — System Design Series
Starts with "System Design Primer: How to start with distributed systems?" and breaks down vertical vs. horizontal scaling from first principles on a whiteboard.
ByteByteGo (Alex Xu)
Watch "7 System Design Concepts Explained in 10 Minutes" and "System Design Interview: A Step-By-Step Guide" for production-grade animations.
The System Design Primer (Donne Martin)
The definitive open-source reference covering Horizontal vs. Vertical scaling, load balancers, caching, sharding, CAP theorem, and mock problems.
-
Phase 1.1: Vertical vs. Horizontal Scaling (Stateless vs. Stateful) IN PROGRESS
-
Phase 1.2: Load Balancing (Round Robin, Least Conn, Consistent Hash) NEXT STEP
-
Phase 1.3: Caching Strategies (Write-through, Cache-aside, LRU/LFU) UPCOMING
-
Phase 1.4 & 1.5: DB Scaling (Sharding, CAP) & Networking (REST, gRPC, WS) UPCOMING
-
Phase 2 & 3: AWS/Azure Cloud Mapping & 45-Min Mock Interview Loop UPCOMING
-
Phase 1 (AI Infra Fundamentals): CPU/GPU/TPU Bottlenecks, Vector DBs (HNSW vs. IVF), Semantic Caching, RAG & Agentic Tool-Calling
-
Phase 2 (Cloud-Native AI Ecosystems): AWS Bedrock vs. Azure OpenAI, OpenSearch/pgvector vs. Azure AI Search, SageMaker vs. Azure ML, Step Functions vs. Semantic Kernel
-
Phase 3 (Modern AI System Design Loop): Token Math, Context Window & KV-Cache Sizing, GPU Latency Budgets & Live AI Support Agent Network Mock
We will unlock Track 2 interactive simulators as soon as we finish Track 1 and you say "Proceed to Track 2".
Test Your Architectural Intuition
Before we move to Step 2 (Load Balancing Algorithms), test this scenario in the simulator or work through it below:
Scenario:
Your e-commerce startup runs on 1 huge Vertically Scaled server. A celebrity tweets about your store, and traffic jumps toward 10,000 RPS. You quickly spin up 3 more identical servers behind a basic round-robin Load Balancer.
Suddenly, angry users report that every time they refresh the page or click "Checkout", they get logged out and their shopping cart disappears! Why did this happen, and how do we fix the architecture so we can scale horizontally without losing user sessions?