How to Virtualise Servers for Cost Savings
Server virtualisation lets you run multiple server workloads on a single physical machine. Most physical servers use only 10-15 percent of their capacity — virtualisation achieves 60-80 percent utilisation, saving money on hardware, power, cooling, and management.
Overview
Virtualisation creates isolated virtual machines (VMs) that each behave like a separate physical server. A single host server with 64GB RAM and fast storage can comfortably run 5-10 virtual servers.
Step 1: Plan Your Virtual Environment
Assess what to virtualise and calculate the resources you need.
Assess Current Server Estate
- List every physical server, its role, and its operating system
- Monitor CPU, RAM, and storage usage on each for at least 2 weeks
- Identify consolidation candidates: Servers using less than 30% of their resources
- Most file servers, print servers, domain controllers, and application servers virtualise well
- Servers that may not virtualise well: GPU-intensive workloads, real-time systems, legacy hardware-dependent apps
- Calculate total resources needed: Sum of all VM requirements plus 20% overhead for the hypervisor
Choose Host Hardware
- Server-grade hardware with error-correcting (ECC) RAM for reliability
- Maximum RAM capacity: 64GB minimum, 128GB+ for larger environments
- Fast storage: Enterprise SSDs or NVMe drives for VM disk images
- Redundant power supplies to prevent single-point-of-failure shutdowns
- Multiple network interfaces for separating management, VM, and storage traffic
- CPU with hardware virtualisation support (Intel VT-x or AMD-V) — standard on all modern server CPUs
- Consider rackmount form factor for data centre environments
Calculate Costs and Savings
- Hardware: One powerful server costs less than 3-5 separate servers
- Power: One server uses far less electricity than multiple servers
- Cooling: Less heat generated means lower air conditioning costs
- Licensing: Some hypervisors are free (Hyper-V, Proxmox, ESXi free tier)
- Management: Fewer physical servers means less maintenance time
- Typical ROI: 3-6 months for a small business virtualisation project
Step 2: Set Up Your Hypervisor
Install and configure the virtualisation platform.
Choose Your Hypervisor
- Hyper-V: Free with Windows Server, easy for existing Microsoft environments
- VMware ESXi: Industry standard, free version available, broadest hardware support
- Proxmox VE: Free and open source, based on KVM/QEMU, excellent web management
- All three support: Live migration, snapshots, high availability, virtual networking
- Hyper-V integrates natively with Active Directory and System Center
- Proxmox has the lowest cost of ownership: Free with optional paid support
Install the Hypervisor
- Download the hypervisor ISO from the official website
- Create a bootable USB drive using Rufus or Etcher
- Boot the server from USB and follow the installation wizard
- Configure management network interface and set a static IP
- Create the initial admin account with a strong password
- Configure storage pools for VM disk images
- Enable remote management access (vSphere Client, Hyper-V Manager, Proxmox Web UI)
Create Virtual Machines
- Allocate CPU cores, RAM, and disk based on each workload's requirements
- Start with conservative allocations — it is easier to add resources than to remove them
- Use thin provisioning for virtual disks: They grow as data is written, saving space
- Install the guest operating system from ISO image
- Install hypervisor guest tools (VMware Tools, Hyper-V Integration Services, QEMU Guest Agent)
- Take a snapshot after initial OS setup — provides an easy rollback point
- Configure virtual networking: Assign VMs to appropriate VLANs
Migrate Existing Servers
- Use P2V (Physical to Virtual) conversion tools: VMware vCenter Converter, Disk2VHD (Hyper-V)
- Or perform a clean install and migrate data and applications manually
- Clean installs are preferred: They avoid carrying over old problems and cruft
- Test the virtualised server thoroughly before decommissioning the physical one
- Keep the old physical server available for 30 days as a fallback
- Update DNS records and IP addresses as needed
Step 3: Ongoing Management and Best Practices
Keep your virtual environment healthy and performant.
Resource Monitoring
- Monitor host CPU, RAM, and storage utilisation — do not overcommit beyond 80%
- Watch for memory ballooning or swapping — indicates insufficient RAM allocation
- Monitor VM disk I/O — slow storage affects all VMs on the same host
- Track VM sprawl: Do not create VMs without a clear purpose and owner
- Use built-in monitoring: vCenter, Hyper-V Performance Monitor, Proxmox dashboard
- Set up alerts for: Host CPU above 85%, RAM above 90%, Storage above 80%
Backup and Snapshots
- Schedule VM-level backups using Veeam (industry standard), Windows Server Backup, or Proxmox Backup Server
- Take snapshots before making changes: Updates, configuration changes, application installs
- Delete old snapshots promptly — they grow over time and degrade performance
- Test backup restoration quarterly to verify backups actually work
- Store backup copies offsite or in the cloud for disaster recovery
- Document backup schedules and retention policies
Maintenance and Updates
- Keep the hypervisor updated with security patches
- Update guest operating systems on a regular schedule
- Plan for host maintenance: Live migrate VMs to another host if available
- Schedule regular reboots of VMs that run 24/7 (monthly is good practice)
- Monitor and replace ageing host hardware before it fails
- Document all VM configurations, resource allocations, and network settings
- Plan for host failure: What happens if your single physical server dies?
If you only have one physical host, ensure you have tested, current backups and a documented procedure to rebuild the entire environment on replacement hardware. Consider a second host for critical workloads.
Need Professional Help?
Our engineers provide expert assistance with setup, troubleshooting, and ongoing support for businesses and individuals across Cornwall.