Introduction

Ballast is a control plane and distributed agent for managing Hyper-V hosts and failover clusters. It stores declarative desired state centrally and runs a lightweight agent on every host that reconciles that state continuously, including while the centre is unreachable. That autonomy is the product's whole thesis: a host keeps honouring its last-known-good configuration when the centre can't be reached, rather than freezing or drifting.

This documentation is organised as:

• Getting Started, what Ballast needs to run, and this page.
• Installation, installing the centre and the agent.
• Your First Cluster, provisioning a host, joining a cluster, or deploying a whole new cluster in one pass with the New Cluster wizard and reusable blueprints.
• Core Concepts, desired-state contracts and quorum witnesses.
• Operations, backup and restore, DR runbooks, and templates.
• Migration, moving workloads off VMware.
• Reference, the full Settings reference, troubleshooting and FAQ.

Who this is for

Ballast is aimed at anyone running Hyper-V standalone hosts or failover clusters who wants a single console for inventory, configuration and recovery, without the licensing model of a VMware-based stack.

Versioning

This documentation is versioned from the 1.0 release onward: the version switcher in the top nav carries a frozen 0.x snapshot (the pre-1.0 docs, no longer updated) alongside 1.0, the current line. Ballast's own grandfathering rule, a deployment keeps the entitlement it was installed under for the life of the release line it's on, is what decided the boundary: 0.x was one ongoing line regardless of minor bumps, so it never needed a per-release snapshot until the line itself changed at 1.0.