How it works

From zero to production-ready private cloud

Step 1

Define

Configure your infrastructure with simple YAML files

Step 2

Validate

CLI performs schema validation and cross-file consistency checks (offline, no SSH)

Step 3

Preflight

Boot each node into rescue mode. The CLI connects over SSH, inspects disks and NICs, computes the per-node disk allocation plan (RAID1 OS, LUKS data, Ceph OSDs) and persists it for apply to consume.

Step 4

Apply

A single command runs the full bootstrap pipeline: wipe and partition disks, install Debian fresh, install Proxmox VE, configure bridges and VLANs, form the cluster, deploy Ceph on top of LUKS-encrypted OSDs.

Step 5

Deploy services

Roll out the workloads tree: global (database, identity, mesh, secrets), regional (bastion, monitoring, SIEM), zonal (loadbalancer, firewall, backup).

Step 6

Operate

Use plan to detect drift between desired and actual state. Add nodes with `soverstack add zone`. Rotate SSH keys, scale flavors, deploy new services anytime.

Deployment timeline

Phase 1

Define & Validate

~5 min

Phase 2

Preflight (rescue inspection)

~5 min

Phase 3

Apply (bootstrap pipeline)

~30 min

Phase 4

Ready for production

~40 min total