Load balancer
Network component that distributes incoming requests across multiple backend servers — provides horizontal scaling, failover and even load distribution.
A load balancer receives incoming connections on a single IP address (VIP) and distributes them by configured algorithms across a pool of 2–N backend servers. Algorithms: round-robin (evenly in turn), least-connections (to the least busy server), IP-hash (the same client always lands on the same backend — important for sticky sessions), weighted (stronger servers get more traffic). Operating layer: layer 4 (TCP/UDP, fast, no TLS intercept) or layer 7 (HTTP-aware, can route paths, terminate TLS, inspect headers). Cloud providers offer managed load balancers as a separate product (Hetzner Cloud LB from €4.90/month, AWS ELB pay-per-use). Self-managed alternatives: HAProxy, nginx, Traefik, Caddy.
Also known as
LB, Traffic distributor
Related terms
Sources
Updated: 16.05.2026