The Bottleneck Problem
Imagine a supermarket with only one checkout lane open. A line forms, customers get angry, and the cashier gets overwhelmed. This is a single server under load. Now imagine opening 10 lanes — but customers are creatures of habit and keep lining up at Lane 1 because it's closest to the door. Lane 1 is crashing while Lane 10 sits empty. The problem isn't capacity; it's distribution.
A Load Balancer (LB) is the employee standing at the front of the store directing customers: "You go to Lane 2, you go to Lane 3." In technical terms, a load balancer sits in front of your backend pool (a group of servers) and distributes incoming client requests across those servers to ensure speed, reliability, and capacity. Clients talk to a single address — the load balancer — and never need to know how many servers are behind it.
Load balancers operate at different layers of the OSI model (Open Systems Interconnection — a conceptual framework that describes how data moves through a network in seven distinct layers). The two layers that matter most for load balancing are Layer 4 (Transport) and Layer 7 (Application). Understanding the difference between them is the foundation of everything else in this lesson.