Scalability in System Design

Scalability in System Design

Horizontal vs Vertical Scaling

Key Concepts

• stateless services — services that hold no session state, enabling horizontal scaling
• load balancer — distributes requests across multiple instances
• read replica — a copy of the database that handles read queries
• write throughput — how many writes the system can handle per second

Useful Phrases

• "To scale horizontally, the service needs to be stateless."
• "We add read replicas to scale read throughput without increasing write load."
• "The database is the bottleneck — we need to either shard it or add a caching layer."