Service · Systems administration

Virtualisation

Plenty of firms arrive at virtualisation because of an invoice: a VMware renewal after the Broadcom licensing changes, or warranties on three physical servers all expiring in the same year. What they actually gain is something else. A virtual server can be copied whole, moved to another host while the first is repaired, or rolled back in minutes to how it was before a failed update. Several ageing boxes that sit mostly idle can be consolidated onto a pair of hosts, where each server is simply a set of files. We compare Hyper-V, VMware and Proxmox with no loyalty to any vendor, three-year cost on the table, and carry out the whole project over a remote console.

3 hypervisors
Hyper-V, VMware, Proxmox
2 hosts
minimum to survive a failure
3-year cost
licences and support included
Tested failover
not merely described

What falls within the scope of this service

We size hosts from real usage data rather than vendor spec sheets, so you neither outgrow them in twelve months nor overpay for idle capacity.

Pin down the details with one of our engineers

Workload measurement

A week of data from your current servers: CPU, memory, disk operations and real peaks, such as ERP month-end or the overnight backup.

Choosing a hypervisor

Hyper-V if you already pay for Windows Server Datacenter, Proxmox if you want to cut licensing and have someone to maintain it, VMware if your backup software or vendor insists. Each option comes with its cost and risk in writing.

Storage

Local NVMe, ZFS, Ceph or a SAN. The right pick follows from how many hosts you run and, most of all, how long you could live without a server after a fault.

Physical to virtual

Each server is converted in a short out-of-hours window. The original stays powered off but untouched for a few days, in case a rollback is needed.

Replication and high availability

Replication to a second host or to a data centre in Spain, and a written procedure setting out what starts first, on which addresses and who informs users.

Whole-machine backups

Hypervisor-level backups that restore an entire server in one step, as well as individual files. With a copy held away from the office.

Templates

Up-to-date Windows Server and Linux images, already monitored. A test server is built from a template in minutes.

How the engagement unfolds, one stage at a time

The timetable depends on how many servers there are and on delivery times for the hardware, which you order from your usual supplier.

01

Measurement

A week of data and a conversation about what is genuinely critical and what can be retired.

02

Design and shopping list

An exact specification to order from your supplier, who delivers and racks it at your office or in the data centre.

03

Remote installation

Once the box is powered and cabled, we log in through the management console, install the hypervisor and networking and benchmark performance.

04

Phased migration

Servers move across one by one, are checked with users, and old hardware is only powered down afterwards.

Two hosts with no shared storage and no replication are not high availability: they are two single points of failure. Buying the second server and considering the matter closed is a common mistake. If the virtual machines live on one host's local disks and that host stops, the other cannot start them. So replication is designed in from the outset, and before signing off the project we deliberately shut down a host to prove the other takes over.

Common questions

Sometimes, for a reason unrelated to saving money: the ability to roll back and bring a machine up on any compatible hardware. If both servers are recent, well backed up and a day of downtime costs you almost nothing, wait until the next refresh. We will tell you if that applies.

The answer is worked out over three years, licences and support included, and it hinges on two often forgotten points: whether your backup software supports the new platform, and who on your team will be able to run it. We migrate an unimportant machine first, let it run for a few weeks and then move the rest.

That is standard nowadays, a3ERP, Sage 200 and SAP Business One included. The condition is that the database gets fast disks and reserved memory. Before migrating we time several routine tasks, such as closing an accounting month, and repeat them afterwards to compare.

Yes, and for many firms that is the best balance: redundant power, cooling and connectivity included, with data kept in Spain or the EU. The comparison is made against the true cost of keeping servers in the office, UPS and power cuts included. The office internet link then becomes the part that needs securing.

No. A snapshot is a return point for a few hours, for instance before an update. Left for months, it grows, slows the machine and ends up filling the storage. Forgotten snapshots have an alert of their own in our monitoring.

Let us work out how many hosts you need

List your current servers and what each one is for. We will reply with a starting configuration and an estimate for licensing.

Hours
Monday to Friday, 9:00-18:00 Spanish time (CET), answers within a working day
Meetings
Video calls via Google Meet or Teams

This site only stores the cookies it needs to work and to remember your chosen city. No advertising or tracking cookies are set. See our privacy policy for the details.