Managed Proxmox · Based in Germany
Infrastructure that survives failure.
We design, operate, and monitor highly available Proxmox clusters for businesses: with automatic failover, redundant storage, round-the-clock monitoring, and a dedicated point of contact.
Services
Our Proxmox clusters combine high availability with redundant Ceph storage – designed, built, and operated from a single source, at a fixed monthly price.
We size your cluster based on load, redundancy needs, and budget: from the virtualization environment through load balancing to the network architecture. You get a design that fits your application, not the other way around.
Learn more →Your cluster is monitored around the clock: utilization, services, hardware, network. We spot anomalies before they become problems and take care of them, without you having to file a ticket.
Learn more →We handle security patches, system updates, and configuration maintenance during live operation, node by node and without maintenance windows for your users. Hardening and a backup strategy are part of the package.
Learn more →Redundant nodes, automatic takeover on failure, regularly tested failover scenarios. A single failure of hardware, network, or software does not bring your application to a halt.
Learn more →When it happens
Not an abstract promise, but a concrete sequence of events. This is how a managed Proxmox HA cluster reacts when a node fails completely:
Your application keeps running while we take care of the cause. You read about it in the incident report – not in complaints from your users.
Architecture principle
High availability does not come from better hardware, but from architecture: every component whose failure would stop your application exists more than once – nodes, load balancers, network paths, storage. Quorum and fencing ensure the cluster makes unambiguous decisions even in a failure scenario, and no split-brain situation can occur.
How this works in detail – from N+1 sizing to regularly tested failure scenarios – is explained on the High Availability Cluster page.
Technology
Virtualization with Proxmox VE and KVM: three-node clusters with quorum, the Proxmox HA Manager for automatic failover, live migration for maintenance without downtime, and rolling updates node by node during live operation – without a full platform outage.
Managed Proxmox Cluster →Distributed Ceph storage across all nodes or ZFS replication for more compact setups: data is stored in multiple copies, a disk defect costs no data. Storage networks are kept separate – and an HA cluster still is no substitute for a backup.
Ceph Cluster →Architecture
A typical managed HA cluster consists of three Proxmox VE nodes joined into a cluster via Corosync. Three nodes are the minimum for a real quorum: if one node fails, the remaining two form the majority and the cluster stays able to make decisions – with no risk of a split-brain situation.
Access from the internet runs through a redundant firewall and load-balancer layer. It distributes requests across the nodes, continuously checks their health, and automatically removes failed systems from the rotation. Below that, the virtual machines run on the nodes; the Proxmox HA Manager monitors them and automatically restarts them on an available target node if a node fails.
The data layer is a distributed storage: depending on requirements, Ceph across all nodes or ZFS with replication. Every data block is stored in multiple copies – the failure of a disk or an entire node costs no data and no operation. For maintenance we use live migration: virtual machines move to another node during live operation, then the freed node is updated and rejoins the cluster.
The result: a highly available Proxmox cluster where hardware failure, network disruption, and maintenance are planned events – not emergencies.
Network
The best failover is useless if the network is the single point of failure. That is why the network layer is designed with consistent redundancy as well:
Behind it stands our own network: AS47263 with fixed IP address ranges, redundant upstreams, and peering in Frankfurt. Details at Network & AS47263.
Process
You describe your project via the contact form, informally and in a few sentences. We get back to you within one business day.
We discuss requirements, load profile, and availability targets. You receive a concrete concept with a transparent fixed-price quote.
We build the cluster, migrate your application, and take over ongoing operations, including monitoring, updates, and a dedicated point of contact.
Operations
All details on ongoing operations can be found at Operations & monitoring.
Why us
Your systems run exclusively in German data centers – primarily in the ISO 27001-certified Maincubes FRA01 in Frankfurt. More: our data centers.
You talk to the person who knows and manages your cluster – no anonymous ticket system, no hold queue.
Data processing under Art. 28 GDPR with a clear agreement. That matters as soon as personal data is stored on the cluster.
Operations contract, SLA, and data processing agreement under German law – with transparent contacts instead of an anonymous corporate structure.
A fixed monthly price for operations and support: predictable and without hidden hourly rates.
Migration
Most cluster projects do not start on a greenfield: there are already virtual machines, standalone hypervisors, or a cluster that has grown over time. We handle the complete migration – from analyzing the existing environment through sizing to a controlled move with minimal, plannable cutover time.
Existing virtual machines move along with network configuration, storage, and backups. Where the environment allows it, we migrate live – without downtime for your users. And just in case, there is always a documented rollback plan.
VMware alternative
Rising license costs after the Broadcom acquisition are turning VMware ESXi into a phase-out model for many businesses. Proxmox VE is the open alternative – without per-socket license costs, with an active community and full control over your own infrastructure.
We migrate your VMware environment to Proxmox, fully managed: analysis, sizing, cluster setup, VM migration, testing, and cutover – including a rollback plan. The complete process is described at Migrating from VMware to Proxmox.
E-commerce
An outage hits high-revenue shops twice: lost orders and lost rankings. Our clusters are built for exactly this load – with scalable app nodes, a replicated database, Redis, and OpenSearch for Shopware and Magento, plus personal support from people who know the shop stack and work directly with your agency.
For shops that must survive even the failure of an entire data center, we build multi-site-capable clusters across Frankfurt and Nuremberg. All details at Cluster hosting for Shopware & Magento.
Data safety
A principle we share with every customer: an HA cluster is no substitute for a backup. Replication protects against hardware failure – but it also replicates mistakes. An accidental deletion, a faulty update, or ransomware lands on all copies immediately.
That is why every managed cluster comes with a separate backup infrastructure: independent backup systems, defined retention periods, and regular restore tests. A backup that has never been restored is just a hope – we turn it into certainty. On request, with an additional copy at a second site.
Reliability
Commitments belong in the contract, not just on the website. In the operations contract we jointly define:
Pricing
Essentials · Ceph HA
€399.99 / month net
The entry into true high availability: three nodes with Ceph, fully managed.
Business · 2 locations
€1,099.99 / month net
Site redundancy included: six nodes across Frankfurt and Nuremberg.
Enterprise
on request
For individual requirements and critical environments.
All prices are net, plus VAT.
All packages come as a fixed monthly price for setup, operations, monitoring, updates, and support – without hidden hourly rates. The exact sizing is defined in the concept; the initial assessment is free and non-binding.
No cluster needed yet? Managed dedicated servers start at €99.99 per month – with the same operations and a clear upgrade path: Managed Hosting. Questions? +49 9382 2209100
FAQ
A group of multiple servers working as one system – designed, built, monitored, and maintained by us. You use the infrastructure; we are responsible for availability, updates, monitoring, and support at a fixed monthly price.
Multiple Proxmox VE nodes are connected via Corosync and form a quorum. The Proxmox HA Manager monitors the virtual machines and automatically restarts them on an available node if a node fails. Details: Managed Proxmox Cluster.
The cluster detects the failure, the HA Manager starts the affected VMs on the remaining nodes, and the load balancer removes the node from the rotation. In parallel, our on-call team is alerted and restores redundancy – your application keeps running in the meantime.
For real high availability, at least three – that way a majority (quorum) remains able to make decisions if one node fails. Two-node setups are possible with an external quorum device, but for business-critical applications we recommend three nodes.
Managed clusters start at €399.99 net per month (Essentials: 3 nodes with Ceph HA storage, fully managed). The Business package with 6 nodes across 2 locations, SLA reports, and monitoring costs €1,099.99 net; Enterprise setups are priced individually. After the free initial consultation, you receive a concept with a fixed monthly price.
Exclusively in German data centers. Processing and storage take place entirely in Germany – GDPR-compliant with a data processing agreement under Art. 28 GDPR.
No. Replication protects against hardware failure but also replicates mistakes: an accidental deletion lands on all copies immediately. That is why every managed cluster includes a separate backup infrastructure with restore tests.
Yes. We take over existing VMs from practically any KVM-based environment, including network and storage configuration. Where possible we migrate live, otherwise with a short, plannable cutover time and a rollback plan.
Yes – it is one of our most common projects. We handle analysis, sizing, setup, migration, testing, and cutover, fully managed. The process is described at VMware to Proxmox.
Yes, for clusters with high demands on redundancy and scaling. For more compact setups we use ZFS with replication instead. Your load profile decides what fits – details at Ceph Cluster.
Rolling: node by node is taken out of the cluster, updated, tested, and rejoined – via live migration, without maintenance windows for your users. Critical security vulnerabilities are patched with priority.
Yes. Hardware, virtualization, services, and network are monitored around the clock, with automatic alerting and on-call duty – at night, on weekends, and on public holidays too.
The cluster takes over automatically: VMs start on the remaining nodes, your application keeps running. Our operations team analyzes the defect, replaces the hardware, and rejoins the node – documented in the incident report.
Yes. After an audit, we connect the environment to our monitoring and take over operations – as-is, or with a step-by-step transition toward the target architecture.
Availability targets, response times by priority level, service hours, and escalation paths are defined in the operations contract – tailored to your application instead of a one-size-fits-all package.
Contact
Briefly describe your application and your requirements. We will get back to you within one business day with an initial assessment, non-binding and free of charge.
Prefer email?
support@cube-hosting.net
Or call us directly:
+49 9382 2209100