Skip to content

Getting Started with Brokkr

This section covers installing and configuring Brokkr.

This is the one list of prerequisites for Brokkr. The other pages link here. Find the path that you will take, and install the tools in its list.

This is Path A of Evaluate Brokkr Locally. It builds Brokkr from source inside Docker and bundles its own Kubernetes (k3s). You do not need a cluster or a Rust toolchain.

  • Docker with Docker Compose
  • Git
  • Angreal, the project’s task runner: pip install angreal
  • curl and jq
  • The brokkr CLI on your PATH, for the agent steps. Download it from the GitHub Release. Each step that uses it also shows the curl form.

This is Path B of Evaluate Brokkr Locally and the Installation guide. These paths use the published images. You do not need a source checkout or a Rust toolchain.

  • A Kubernetes cluster, v1.29 or later (the agent chart declares kubeVersion: ">=1.29.0-0"). For an evaluation, a local kind or k3d cluster is sufficient.
  • kubectl, configured to reach that cluster
  • Helm 3.8 or later (installation guide)
  • Docker, to run kind or k3d, and to run brokkr-broker generate-pak from the published image
  • curl and jq
  • The brokkr CLI on your PATH, for the agent steps. Download it from the GitHub Release. Each step that uses it also shows the curl form.

This is the Local Development Environment. It builds and tests Brokkr from source.

  • Rust 1.90 or later (the workspace uses edition 2024)
  • PostgreSQL client tools: libpq, which the build links, and psql. The database itself runs in Docker.
  • Docker with Docker Compose
  • Git
  • Angreal, the project’s task runner: pip install angreal
  1. Evaluate Brokkr Locally — get a working Brokkr in front of you fast
  2. Installation — install Brokkr on your system
  3. Configuration — configure Brokkr for your environment
  4. Local Development Environment — run the whole stack from source, for contributors

Whichever path you take, the broker also serves a read-only Operator Console at its own root URL once it is running — open it in a browser to see your fleet, deployments, and telemetry without setting anything up.

After completing the getting started guide, you can: