DOCUMENTATION / GETTING STARTED

Your first route.

From a single binary to a healthy backend pool. Start locally, understand the flow, then make it production-ready.

01. Install the binary

OpenLoadBalancer runs on Linux, macOS, and Windows. The upstream installers select a release binary for your platform. Review the shell script or PowerShell script before running it.

Linux / macOS
curl -fsSL https://openloadbalancer.dev/install.sh | sh
Windows · PowerShell
irm https://openloadbalancer.dev/install.ps1 | iex

Prefer a manual installation? Download the matching binary from GitHub Releases and add it to your PATH. Homebrew users can run:

Homebrew
brew tap openloadbalancer/olb
brew install olb

02. Define a backend pool

Save the configuration below as olb.yaml. It listens on port 8080 and distributes traffic between two local services. Replace the example addresses with reachable backends that return a successful response on /health.

olb.yaml
version: 1 admin:  address: "127.0.0.1:9090" listeners:  - name: http    protocol: http    address: ":8080"    routes:      - name: default        path: /        pool: backend pools:  - name: backend    algorithm: round_robin    health_check:      type: http      path: /health      interval: 10s      timeout: 5s    backends:      - id: backend-1        address: "127.0.0.1:7091"      - id: backend-2        address: "127.0.0.1:7092"

The admin API binds to 127.0.0.1:9090. Keep it on a trusted interface; the public listener and admin listener serve different purposes.

The configuration structure follows the product repository’s minimal configuration. YAML, JSON, TOML, and HCL formats are supported, including environment-variable substitution. Use olb setup for the interactive configuration wizard.

03. Validate and start

Validate & start
olb config validate --config olb.yaml
olb start --config olb.yaml

With your backend services running, make a request through the listener:

Check the listener
curl -i http://localhost:8080/

Visit http://127.0.0.1:9090 on the same machine for the admin interface. If a backend is unavailable, health checks remove it from rotation until it recovers.

04. Keep a clear view

The CLI exposes operational commands without requiring you to build another dashboard. Use the upstream API reference for programmatic integrations.

Daily operations
olb status --api-addr localhost:9090
olb backend list --api-addr localhost:9090
olb health show --api-addr localhost:9090
olb reload --api-addr localhost:9090

For automated monitoring, connect Prometheus metrics and structured JSON logs. The built-in MCP server lets compatible AI tools inspect metrics, diagnose issues, and manage backends. Configure authentication and audit logging before enabling management access.

05. Before production

  • Configure TLS termination, ACME certificates, or mTLS for your environment.
  • Keep admin and MCP interfaces restricted to trusted networks and configure authentication.
  • Set appropriate health-check intervals, timeouts, and request limits.
  • Choose a balancing algorithm that matches your workload and test with your traffic.
  • Review observability, graceful draining, and clustering in the production deployment guide.

Published benchmark results describe the upstream test environment. Measure latency, throughput, and resource use on your own hardware before capacity planning.

Continue exploring