Introduction to HAProxy
Every application has the same unglamorous requirement: stay up. Traffic doesn't arrive politely, and servers don't fail on schedule. A single server can't absorb that reality. A load balancer can, and the most widely used software load balancer is HAProxy.
HAProxy sits between your users and your servers. It spreads requests across healthy backends, detects the ones that stop responding, and routes around them before anyone notices. It terminates TLS, enforces routing rules, and exposes detailed stats on every connection it handles. All of it runs from a single binary and one readable config file.
In this session, you'll learn:
The problems load balancing solves, from failed servers to zero-downtime deploys
How a request travels through HAProxy, from client connection to backend response
The four sections of a basic haproxy.cfg (global, defaults, frontend, backend) and what each one controls
How HAProxy decides where each request goes, and why the default choice isn't always the right one
Where HAProxy fits in a modern stack, whether you run bare metal, VMs, containers, or Kubernetes
This session is built for people who are new to HAProxy, whether or not they are new to load balancing. No HAProxy experience required. You'll leave with a working mental model of the software and a better answer to one question that trips up a lot of newcomers: which load balancing algorithm to pick. The HAProxy Config series picks it up from there.
Speakers
