System Design
Load Balancer
A component that spreads incoming requests across multiple servers instead of sending them all to one.
Reviewed by the RadarTrek editorial team · June 2026
A load balancer sits in front of a pool of servers and distributes each incoming request to one of them — by round robin, by least connections, or by other strategies — so no single server gets overwhelmed. It's also what makes horizontal scaling possible: add another server behind the load balancer, and it immediately starts taking traffic.
Why it matters
- —A load balancer is what turns "one server" into "a pool of servers" without the client ever knowing the difference.
- —It only works cleanly if servers are stateless — anything stored in one server's local memory won't be there on the next request.
- —Health checks let a load balancer stop sending traffic to a server that's failing, before users notice.
Where to learn this
🎓
Load Balancing and Horizontal Scaling
System Design for Builders course
This is the exact lesson that covers this term in depth — with examples, diagrams, and a hands-on exercise.