XCP-ng + Xen Orchestra
Fully open-source Xen hypervisor plus unified management and backup, sold by Vates with simple per-host pricing — the other credible open-source VMware exit.
How we testXCP-ng is the open-source Xen-based hypervisor; Xen Orchestra (XO) is the web management, backup and DR layer on top. Vates (France) sells the two as one supported stack, priced per host rather than per core or socket. XO 6.x has invested heavily in the VMware V2V path (direct ESXi/vCenter import with delta migration, 16 TB disks since 2026), and the automation story is unusually complete for an open-source vendor: a maintained Terraform provider, Packer, Ansible, REST API with RBAC, Go SDK and a Kubernetes CSI driver. The self-service portal offers resource sets with quotas but no chargeback or billing.
Pros
- +Transparent per-host pricing with backup/DR included — no per-core minimums
- +Strong, fast-improving V2V migration from VMware
- +Full open-source stack with a coherent automation/IaC toolchain
Cons
- −Xen ecosystem and ISV certification narrower than KVM or ESXi
- −Self-service is quota-based only — no metering or billing
- −Smaller vendor; some limits (large disks, formats) only recently lifted
Pricing
Essential
$2,000/yr · annual commitment
Month-to-month: not published
- ✓Up to 3 hosts, 6 tickets, business-day support
Essential+
$4,000/yr · annual commitment
Month-to-month: not published
- ✓Up to 3 hosts, unlimited tickets, all features
Pro
$1,000/yr · annual commitment
Month-to-month: not published
- ✓Per host, min 3 hosts, business-day
Enterprise
$1,800/yr · annual commitment
Month-to-month: not published
- ✓Per host, min 4 hosts, 24/7, 1h critical response
Specifications
| Rating | 4.0 |
| Licensing | Per host / node |
| Backends | Xen (XCP-ng), Citrix XenServer |
| Self-service | ✓ |
| Metering / billing | No |
| VMware migration | ✓ |
| Multi-backend | No |
| IaC | Terraform / OpenTofu (official), Packer, Ansible +2 |
Researched and drafted with AI assistance; reviewed and approved by Tim Meredith. How we use AI