Architecture Overview
This document provides an overview of the architecture used in the testbed for
kubevirt-actions-runner.
The runner pod receives GitHub Actions metadata, creates a VirtualMachineInstance (VMI) from a predefined KubeVirt template, waits until the job reaches a terminal state, and then removes runtime resources.
Components
- kind: Creates a local Kubernetes cluster running in Docker.
- KubeVirt: Enables VM workloads inside Kubernetes.
- VM Template: Cloned dynamically by the runner.
- Demo Workload: Validates VM lifecycle management.
Deployment Flow
The following diagram describes the complete deployment, runtime execution, and teardown sequence.

Related documentation
- For a guided first run, see the Quickstart tutorial.
- For operational setup tasks, see How-to guides.
- For command and configuration details, see Reference.