HA virtualisation on UK metal — Proxmox VE + Ceph
Production-grade Proxmox VE clusters with hyper-converged Ceph storage, live migration and Proxmox Backup Server — running on HPE ProLiant nodes in our privately owned Cambridge Tier III data centre. A real alternative to VMware vSphere, without per-core licensing.
Live migration
Zero downtime
Hyper-converged
Ceph storage
3+ nodes
True quorum & HA
Live migrate
Zero-downtime moves
PB-scale
Hyper-converged Ceph
24/7
UK NOC operations
Cluster shapes we build
Three reference architectures — every cluster is sized to your VM count, IOPS, durability and growth targets.
Starter Cluster
3-node entry cluster — quorum from day one
From £649/mo
- Nodes: 3 × HPE ProLiant DL360 Gen11
- CPU: Intel Xeon Silver 4416+ (20C / 40T) per node
- RAM: 128 GB DDR5 ECC per node
- Storage: 2 × 3.84 TB NVMe per node (Ceph 3x)
- Network: 2 × 10 GbE public + 2 × 10 GbE cluster
- Proxmox VE + Ceph + PBS, fully managed
Production Cluster
Most popular — HA + Ceph at sensible scale
From £1,499/mo
- Nodes: 4–6 × HPE ProLiant DL360 Gen11
- CPU: 2 × Intel Xeon Gold 6430 per node (64C / 128T)
- RAM: 256 GB DDR5 ECC per node
- Storage: 4 × 7.68 TB NVMe per node (Ceph 3x or EC)
- Network: 2 × 25 GbE public + 2 × 25 GbE cluster (LACP)
- Proxmox VE + Ceph + PBS, fully managed
Enterprise Cluster
Dual-socket DL380 with hyper-converged Ceph
Built to order
- Nodes: 6+ × HPE ProLiant DL380 Gen11
- CPU: 2 × Intel Xeon Platinum 8480+ per node
- RAM: 512 GB – 4 TB DDR5 ECC per node
- Storage: Hybrid NVMe + HDD Ceph pools, PB-scale
- Network: 2 × 100 GbE public + 2 × 100 GbE cluster
- Proxmox VE + Ceph + PBS, fully managed
Prices shown ex-VAT. Larger clusters, GPU passthrough, dedicated Ceph tiers and off-site PBS replication available on request.
Everything a serious private cloud needs
HA, hyper-converged storage, live migration, backups and SDN — engineered together, operated by our UK NOC.
True HA Clustering
Corosync-quorate Proxmox VE clusters with automatic VM failover — lose a node and your guests restart on a healthy host in seconds.
Hyper-Converged Ceph
Ceph RBD on the same nodes — replicated or erasure-coded pools, self-healing, no SAN to fail. Block, file and S3 from one fabric.
Live Migration
Move running VMs between nodes with zero downtime for maintenance, balancing or hardware refreshes — over a dedicated 25 / 100 GbE migration network.
VMs and LXC Containers
Run full KVM virtual machines and lightweight LXC containers side-by-side, managed from one Proxmox web UI and API.
Proxmox Backup Server
Integrated, deduplicated, incremental backups to a dedicated PBS appliance — instant file-level restore and verifiable backup chains.
SDN & VLAN Networking
Proxmox SDN with VLANs, VXLAN and EVPN. Per-tenant networks, private inter-VM fabrics and integration with our AS48825 edge.
600 Gbps DDoS Included
Always-on, in-line Corero SmartWall mitigation on every public IP. No add-on fees, no opt-in.
Managed by UK NOC
We run the cluster — Proxmox VE upgrades, Ceph health, PBS verification, kernel patching, capacity planning — handled 24/7 from Cambridge.
What's running on the cluster
A complete, integrated stack — installed, hardened and supported.
| Component | Role | Notes |
|---|---|---|
| Proxmox VE | Hypervisor & cluster manager | KVM + LXC, HA groups, replication, SDN, web UI & API |
| Ceph (RBD / CephFS / RGW) | Hyper-converged storage | Replicated or erasure-coded, self-healing, scales online |
| Proxmox Backup Server | Backup & restore | Incremental forever, dedup, encryption, verify jobs |
| Corosync / Pacemaker | Cluster quorum & HA | Sub-second failure detection, automatic VM restart |
| Proxmox SDN | Software-defined networking | VLAN, VXLAN, EVPN, per-tenant isolation |
| Cloud-init / Terraform | Provisioning & IaC | Templates, golden images, automated VM rollouts |
Who we run Proxmox clusters for
From private clouds to VMware migrations — same UK metal, same UK NOC.
Private clouds
Drop-in alternative to VMware vSphere — same HA, live migration and shared storage, without per-core licensing or hyperscaler bills.
Multi-tenant platforms
MSPs and agencies running isolated customer VMs with per-tenant networks, storage pools and quota — all behind one managed cluster.
Compliance & data sovereignty
UK-only Proxmox + Ceph clusters for finance, healthcare, legal and public-sector workloads — ISO-aligned operations, no overseas replicas.
VMware migration
Tired of Broadcom pricing? We migrate ESXi and vSphere estates to Proxmox + Ceph with planned, low-downtime cutovers and full rollback.
A proper reference architecture
Every cluster is engineered to your VM count, IOPS, durability and growth targets — not pulled off a template. Separate cluster, migration and storage networks; failure-domain-aware Ceph placement; verified PBS backups; tested failover from day one.
- Three networks, not one — public, Corosync cluster and Ceph/migration on separate 10/25/100 GbE pairs for predictable performance under load.
- CRUSH-aware Ceph — placement across hosts and racks so the cluster survives real-world failures, not just disk loss.
- Tested HA & restore — we pull a node and restore a VM from PBS before handover. You see it work before you sign off.
- Monitored 24/7 — cluster quorum, Ceph PG health, PBS verify status, version drift and capacity — handled by our NOC.
Example 5-node production cluster
Nodes : 5 x HPE ProLiant DL360 Gen11
Per node : 2 x Xeon Gold 6430 (64C / 128T)
256 GB DDR5 ECC
4 x 7.68 TB NVMe (Ceph OSD)
2 x 480 GB NVMe (OS, mirrored)
Networking : 2 x 25 GbE public (LACP, VLANs)
2 x 25 GbE cluster + migration (LACP)
Hypervisor : Proxmox VE 8.x, HA enabled
Storage : Hyper-converged Ceph, 3x replica
~80 TB usable VM storage
Backups : Dedicated Proxmox Backup Server
7-day incremental, verified daily
DDoS : 600 Gbps Corero SmartWall, always-on
Operations : Fast2host UK NOC, 24/7Configurations are illustrative — every cluster is sized to your workload.
How to get started
01
Design the cluster
Tell us VM count, vCPU / RAM totals, storage capacity and durability goals. We size nodes, choose replica vs erasure coding, plan the SDN topology and the right migration network.
02
Build & bootstrap
We rack the nodes in Cambridge, cable separate cluster and migration networks, install Proxmox VE and Ceph, configure HA groups, PBS and SDN, then run failover and recovery tests before handover.
03
Operate & scale
We monitor cluster, Ceph, PBS and SDN 24/7 — handling upgrades, OSD replacement, capacity expansion and incident response. Add nodes online; the cluster rebalances itself.
8 common questions