Where the group runs, and what is live.
The group is built on owned and rented compute, run in-house. Customer workloads run in the United States today, a European core node handles management for everything, and new regions come online as customers in them appear.
Global presence
North America and Europe live, UK and Asia mapped
Customer-facing compute and the management core today, with the next two regions called out honestly as planned and greyed until a node exists.
United States production compute. Customer workloads run here today, across paired locations.
A European core node and a large dedicated machine. Orchestration, billing, monitoring and backups for the whole group.
A UK region is mapped and ready to stand up. It opens once demand there warrants its own node.
An APAC region is planned. It follows once customers in the region ask for it, not before.
Live regions have a node running today. Planned regions are mapped and greyed until a node is stood up there, so the map never claims a presence we do not have.
Two live regions
North American production compute and a European core node run today. Customer workloads sit in North America; management and the control plane sit in Europe.
Two mapped regions
The United Kingdom and Asia-Pacific are designed into the network and ready to stand up. They open on real demand, not ahead of it.
Build and infrastructure status
What is operational, what is in build
A plain status board for the shared platform. We would rather show the in-build items than hide them.
| Component | What it does | Status |
|---|---|---|
| Production cluster | Multi-region compute cluster carrying live workloads | Operational |
| Control plane | Virtualization and orchestration across the nodes | Operational |
| Billing and provisioning | Customer billing and automated provisioning at Overnight.Host | Operational |
| Monitoring and alerting | An independent monitor pushing alerts off-platform | Operational |
| Public status page | Live service status at up.overnight.host | Operational |
| Off-site backup | Object-lock and storage-box backup beyond the core node | In build |
| Dedicated automation node | A separate node for the automation line | Planned |
| United Kingdom and Asia-Pacific regions | Mapped into the network, stood up on real demand | Planned |
How we scale
Proof first, then metal
We start small on rented infrastructure and grow onto more, and eventually owned, hardware as real usage demands it.
Rented seed
Today the cluster is rented. It is the seed, not the ceiling. New nodes are ordered against real commitment, not hope.
Region on demand
A new region opens when customers in it appear, so capacity tracks demand instead of running ahead of it.
Isolation and sovereignty
Workloads are isolated by business, and the automation and hosting lines are positioned on European data sovereignty.