Solution
Managed Proxmox Cluster
Three nodes with quorum, HA Manager and Ceph storage, fully operated.
Product · Dedicated servers
In brief
A managed dedicated server is a physical server reserved for you that we run end to end: operating system, services, configuration, hardening, updates, and 24/7 monitoring and alerting. On an unmanaged server, those tasks stay with you. Our managed dedicated servers in Germany run on AMD Ryzen with 64 GB RAM and NVMe storage, from €99.99 net per month.
Models and pricing
Our managed dedicated servers come in three models: Starter Server at €99.99, Business Server at €149.99 and Pro Server at €199.99 per month, net. All three include 64 GB RAM, two 1 TB NVMe drives, IPv4 and an IPv6 /64 network, plus DDoS protection and monitoring. They differ in processor and uplink; full operations are included with every model.
| Specification | Starter Server | Business Server | Pro Server |
|---|---|---|---|
| Processor | AMD Ryzen 7 5700X | AMD Ryzen 9 5900X | AMD Ryzen 9 7900 |
| Memory | 64 GB RAM on every model | ||
| Storage | 2× 1 TB NVMe on every model | ||
| Network uplink | 2× 1 Gbit/s (LACP) | 2× 1 Gbit/s (LACP) | 10 Gbit/s |
| IP addresses | IPv4 + IPv6 /64 on every model | ||
| Protection and monitoring | DDoS protection and monitoring on every model | ||
| Operations | fully managed on every model | ||
| Price per month, net | €99.99 | €149.99 | €199.99 |
Every server comes at a fixed monthly price including operations, with no hidden hourly rates. All prices are net, plus VAT.
Every model is a managed bare metal server: the processor, memory and NVMe storage belong to your server alone, with no shared resources. In terms of specifications, Starter and Business differ only in the processor. Pro is the only model with a 10 Gbit/s uplink.
We put together other hardware configurations on request. Cluster packages and servers are listed side by side on our complete pricing page.
Scope of services
With a managed server, we take over everything beneath your application: the operating system, services, configuration, hardening, updates, and 24/7 monitoring and alerting. It is managed hosting in Germany for companies that need dedicated performance and would rather not build their own server operations. You look after your application, we look after everything beneath it.
| Task | Unmanaged server | Managed server with us |
|---|---|---|
| Operating system and services | You | We setup and operation |
| Configuration and hardening | You | We changes to the system configuration agreed with you |
| Updates and security patches | You | We critical vulnerabilities first |
| Monitoring and alerting | You | We around the clock, including nights and weekends |
| Backups | You | On request separate backups with restore tests |
| Your application | You | You we run everything beneath it |
How we apply security patches and harden systems is described on our page about managed patch management and hardening.
Decision guide
A managed server suffices if your application can tolerate downtime until a hardware fault is fixed. If a hardware fault must not stop your service until it is repaired, you need a managed Proxmox cluster, where the HA Manager automatically restarts affected virtual machines on another node. The Business package spreads six nodes across FRA01 and NBG01.
| Criterion | Managed server | Managed Proxmox cluster (Essentials) | Business package (2 sites) |
|---|---|---|---|
| Setup | one dedicated server | 3 nodes with quorum | 6 nodes across FRA01 and NBG01 |
| Storage | 2× 1 TB NVMe in the server | Ceph, every data block stored three times on different nodes | Ceph HA storage replicated across both sites |
| Failure behavior | No failover If the server fails, the application is unavailable until the fault is fixed; 24/7 monitoring alerts us. | Automatic If a node fails, the HA Manager restarts the affected VMs on another node. Services running on several nodes behind the load balancer remain available. | If a site fails: a documented switchover plan, tested regularly |
| Maintenance and updates | Updates and security patches on the single system; a reboot interrupts the service because no second node takes over. | Rolling updates node by node with live migration while your systems keep running. Exceptions such as a major database upgrade are announced in advance. | rolling updates as in Essentials |
| Monitoring and reporting | 24/7 monitoring and alerting | 24/7 monitoring and alerting, monthly operations report | adds SLA reports and extended monitoring with escalation |
| Price per month, net | from €99.99 (Starter Server) | €449.99 | €1,099.99 |
| A good fit if | an outage until the fault is fixed is acceptable | a hardware fault must not stop the service until it is repaired | your requirements call for a second site |
All prices are net, plus VAT.
The difference is redundancy. A single server is managed and monitored, but there is no second node to take over. In a cluster, the load balancer removes a failed node from rotation within seconds. The node's virtual machines restart on another node after fencing; how long that takes depends on fencing and on the guest system's boot time.
Our page on high availability clusters explains step by step how automatic failover in an HA cluster works and which layers are built redundantly.
Upgrade path
When your availability requirements grow, you move from a managed server to a managed Proxmox cluster. We build the cluster in parallel, test failover and migrate your application with a rollback plan. Your server keeps running until the switchover, and you choose the date.
Phase 01
We document your application, its load profile and your availability and recovery targets: the tolerable downtime (RTO, recovery time objective) and the tolerable data loss (RPO, recovery point objective).
Phase 02
You receive a written design with an architecture diagram, a component list and a fixed price for build and operations, usually within a few business days.
Phase 03
The cluster is built alongside your server, including monitoring and backups. Before handover we run a full failover test. Build and testing typically take one to three weeks.
Phase 04
We migrate your application following a migration plan with minimal, predictable switchover time, on the date you set. If problems occur, the rollback plan takes effect.
Phase 05
From then on we operate the cluster: 24/7 monitoring with on-call duty, rolling updates and a monthly operations report covering availability, incidents and utilization.
The target platform is the managed Proxmox cluster with at least three nodes and quorum via Corosync. In the Essentials and Business packages it uses Ceph storage, which keeps every data block three times on different nodes. How assessment, sizing and the failover test fit together in a project is covered under cluster design and setup.
FAQ
Who runs it. With an unmanaged server, often sold in Germany as a root server, you administer the system yourself: updates, hardening, monitoring and responding to alerts are your job. With a managed server we take over the operating system, services, configuration, hardening, updates, and round-the-clock monitoring and alerting. A fully managed server leaves you free to focus on your application.
It depends on how long your application may be down. If it can tolerate an outage until a hardware fault is repaired, a managed server is enough. If it has to survive the failure of a server, you need a cluster in which the Proxmox HA Manager restarts affected virtual machines on another node. Our availability calculator shows how much downtime per year an availability target allows.
On request. For managed servers, we set up backups when you ask for them, as separate backups with restore tests, matched to your data. Why only a restored backup counts as verified is explained on our page about backup and disaster recovery.
Yes. We build a managed Proxmox cluster in parallel to your server, test failover before handover and migrate your application with a rollback plan. You set the switchover date. The upgrade path from server to cluster lists each step.
Yes, on request. The Starter, Business and Pro models cover the standard specifications; we put together other hardware configurations on request. Tell us what you need in terms of processor, memory, storage and uplink.
Yes, by agreement. You get visibility into your server environment and access arranged with you. We plan changes to the system configuration together so that stable operations remain our responsibility.
From €99.99 per month, net, for the Starter Server with an AMD Ryzen 7 5700X. The Business Server with a Ryzen 9 5900X costs €149.99 and the Pro Server with a Ryzen 9 7900 and a 10 Gbit/s uplink €199.99, each per month, net, with full operations included.
Next step
Describe your application, the expected load and the longest outage you can accept. We recommend the right model or tell you when a managed Proxmox cluster is the better choice.
Request a managed serverRelated