Skip to content
Infraexa

Cluster 3

Cluster 3 is three Core-class nodes, each with 48 cores, 128 GB of memory and two 1.92 TB NVMe drives, joined by a private network and delivered as one system: a highly available Kubernetes control plane, a replicated Postgres with automatic failover, shared ingress and certificates, and a GitOps pipeline wired to your repositories. One node can fail without anyone noticing.

Request a quote for Cluster 3Compare with other packages

No setup fee. Monthly billing, three-month minimum, then cancel with 30 days notice. Delivery within 48 hours for single servers; about a week for clusters.

Cluster 3
Cluster series
Processor
3 × AMD EPYC 9454P
Cores
144 cores / 288 threads total
Memory
384 GB DDR5 ECC total
Storage
6 × 1.92 TB NVMe total
Network
3 × 10 Gbit, private VLAN, 60 TB included
Response
Engineer within 15 min
€9,850
per month, excluding VAT, everything included
Request a quote

Datasheet

Exact configuration as delivered. Variants within a series are quoted individually.

Nodes
3 × AMD EPYC 9454P, 48 cores / 96 threads each
Memory
128 GB DDR5 ECC registered per node, 384 GB total
Storage
2 × 1.92 TB NVMe SSD per node, mirrored; replicated persistent volumes across nodes
Network
10 Gbit uplink per node, private VLAN between nodes, 60 TB outbound per month included
Kubernetes
Three-node highly available control plane, ingress, cert-manager, external-dns, metrics and log shipping preinstalled
Database
Replicated Postgres with automatic failover and point-in-time recovery; Redis Sentinel on request
Delivery
GitOps (Argo CD or Flux) connected to your repositories; staging and production namespaces
Location
All nodes in one site in Germany or Finland; a cross-site variant is quoted separately
Backups
Nightly encrypted snapshots of volumes and database to a second site, 30-day retention
Incident response
Critical issues answered by an engineer within 15 minutes, around the clock; named engineer and fortnightly review

What Cluster 3 is for.

Workloads this configuration has been sized for, and what it adds beyond the standard layer.

Best for

  • Production SaaS that needs to survive a node failure
  • Teams replacing a managed Kubernetes bill plus managed database bill
  • Agencies hosting many client applications on one platform
  • Regulated products that need an EU-only control plane

Included beyond the standard layer

  • Highly available Kubernetes with GitOps
  • Replicated Postgres with automatic failover
  • Fortnightly review with your named engineer

Included with every package.

Nothing on this list is an upsell.

Provisioning and hardening within 48 hours
Managed operating system
Observability
Backups that restore
Security
Network
Engineers around the clock
Exa, the AI operations engineer

Solutions built on this package.

Guides for this kind of server.

Buying9 min

Dedicated servers vs cloud for a SaaS with a €20,000 monthly bill

A worked migration of a typical SaaS from hyperscaler instances, managed database and egress to a few dedicated servers, with the before-and-after numbers.

Read the guide
Operations9 min

Sizing a dedicated server for Postgres: cores, memory, NVMe

How to pick memory, storage and CPU for production Postgres on dedicated hardware, and why one 48-core NVMe server replaces a lot of managed database spend.

Read the guide
Operations10 min

Kubernetes on bare metal: a three-node high-availability design

How to build a production Kubernetes cluster on three dedicated servers: control plane, storage, networking, ingress and the operations that keep it running.

Read the guide

Other packages.

Ready to run Cluster 3? Tell us about the workload.

A quote within one business day. €9,850 per month with everything included, no setup fee.

Request a quoteTalk to an engineer