Load Balancers

A load balancer sits between clients and your server fleet. It routes incoming requests, detects unhealthy servers, and enables horizontal scaling. Without one, you have a single point of failure.

What a load balancer does

Core responsibilities at the network layer.

• ›Distributes requests across multiple servers
• ›Health checks servers — stops routing to failing nodes
• ›Enables rolling deploys without downtime
• ›Provides a single public IP for a fleet of private servers
• ›Can terminate SSL, handle rate limiting, or add headers

Routing algorithms

How requests get assigned to servers.

• ›Round robin — each server in turn. Simple, works for stateless services.
• ›Least connections — route to the server with fewest active connections.
• ›IP hash — same client IP always hits the same server. Useful for session stickiness.
• ›Weighted round robin — newer or more powerful servers get more requests.

L4 vs L7 load balancing

The layer determines what decisions the balancer can make.

• ›L4 (transport) — routes based on IP + TCP port. Fast, but no content awareness.
• ›L7 (application) — routes based on HTTP headers, URL, or body. More flexible, slightly more overhead.
• ›Use L7 when you need path-based routing, A/B testing, or content inspection.

Interview tips

• ✓Mention load balancers early in any high-traffic design
• ✓Specify stateless vs stateful — it determines your routing strategy
• ✓Always address: what if the load balancer itself fails?
• ✓In cloud designs, say 'AWS ALB' or 'NGINX' to show real-world familiarity

Follow-up questions to expect

• ?What happens if the load balancer goes down?
• ?How do you handle session state when requests can hit any server?
• ?When would you use a L4 vs L7 load balancer?