Home / All services / Infrastructure
Virtualisation & clusters
The real question isn't how many servers you have. It's what happens when one of them stops this afternoon.
Most mid-sized businesses run everything on one or two physical servers. It works until it doesn't, and then recovery is measured in days because the replacement part has to be ordered. A cluster isn't a luxury; it is how you make that day boring.
What we build
- A cluster of two or more nodes on shared storage.
- Roles separated into their own virtual machines rather than stacked on one.
- Live migration of machines between nodes, with no service interruption.
- Monitoring that warns before a disk fills, not after.
- Full documentation: what runs where, and what falls over if it reboots.
Migrating what you already have
Existing servers can usually be converted into virtual machines without rebuilding everything. We do it out of hours, with the old system ready to come back up if anything goes wrong. In practice the planned downtime stays under an hour; often there is none at all.
Sizing
We don't sell more hardware than you need. Before quoting, we measure the real load — CPU, RAM, IOPS — over a period that includes month end. Then we add three years of growth, so you don't buy twice.
Not sure where to start?
The audit puts the whole picture on paper within two weeks, at a fixed price. The report stays yours whatever you decide afterwards.