Skip to content

Virtualization

Superphenix provides virtual machines as a core feature: you run VMs on the same platform that runs the control plane, with the same declarative model and automation as the rest of the stack.

What you get

  • VMs as first-class workloads — Create, start, stop, and manage VMs via the same APIs and GitOps workflows you use for the rest of the platform. No separate hypervisor management layer.
  • Live migration — Move VMs between nodes with minimal downtime for maintenance or load balancing.
  • Snapshots and clones — Snapshot running VMs and clone them for backups, templates, or duplicate environments.
  • Templates and images — Boot VMs from templates or images (e.g. from a registry). Inject users, SSH keys, and network config at first boot.
  • Persistent and ephemeral disks — Attach persistent block storage to VMs, or use ephemeral disks for stateless or test workloads.
  • VM pools — Run replica sets of VMs (e.g. for Kubernetes-as-a-Service node pools) with the same ease as scaling other workloads.
  • Node drain and placement — Draining a node migrates VMs off it. Use affinities and taints to control where VMs run.

Access to VMs

  • SSH — VMs can be configured with SSH keys at boot and accessed over the network.
  • Serial console — Direct serial console access when SSH is not available.
  • VNC — Graphical console access (e.g. for Windows VMs), with access control and auditing via the web console.

Integration with the rest of the platform

  • Networking — Each VM can be attached to one or more networks (VPCs and subnets). IPs are stable across restarts and live migration.
  • Storage — VM disks are backed by the platform’s block storage, with snapshot and clone support and optional replication across AZs.

See Network and Storage for those capabilities.