Reviewing your virtualization platform? How to evaluate hypervisor options on requirements, total cost and migration risk before you switch.
Virtualization has been the backbone of enterprise data centres for two decades. Changes in licensing and packaging across the market have prompted many organisations to review whether their current platform is still the right fit.
Start with requirements, not products
- How many hosts, virtual machines and sites do you run?
- Which features do you rely on — live migration, high availability, distributed switching, replication?
- Which backup, monitoring and security tools must integrate?
- What skills does your team have today?
The main options
Choices broadly include commercial enterprise hypervisors, Hyper-V within the Microsoft ecosystem, KVM-based platforms (including open-source and commercially supported distributions), hyperconverged platforms with an integrated hypervisor, and moving selected workloads to public cloud or containers.
Evaluate total cost, not licence price
Include licences or subscriptions, support, hardware compatibility, training, migration effort and any changes to backup and disaster recovery tooling. A cheaper licence can be outweighed by migration and retraining costs.
Plan migration carefully
- Inventory VMs and dependencies; retire what is no longer needed.
- Pilot with non-critical workloads first.
- Validate performance, backup and recovery on the new platform.
- Migrate in waves with clear rollback plans.
Tip: a platform review is a good moment to decide which workloads should be containerised or moved to SaaS instead of migrated as-is.
5 best criteria for comparing hypervisor options
- Workload compatibility. Confirm that critical applications, guest operating systems and appliances are supported and certified.
- Operational features. Compare live migration, high availability, resource scheduling, snapshots and backup integration.
- Ecosystem support. Check integration with your backup, monitoring, security and automation tools.
- Total cost of ownership. Include licences or subscriptions, support, hardware requirements, training and migration effort over several years.
- Vendor and community health. Assess roadmap clarity, support quality and, for open-source options, the strength of the community and commercial backers.
Planning a migration
Run a proof of concept with representative workloads, including databases and latency-sensitive applications. Use migration tools to convert virtual machines, test performance and failover, and move workloads in waves with clear rollback plans.
Alternatives to consider
- Moving suitable workloads to public cloud infrastructure.
- Adopting hyperconverged platforms with integrated virtualisation.
- Containerising applications and running them on Kubernetes, including platforms that manage virtual machines alongside containers.
Common mistakes to avoid
- Choosing based on licence cost alone and underestimating migration effort.
- Forgetting third-party appliances that only support specific hypervisors.
- Skipping staff training on the new platform.
Frequently asked questions
How long does a hypervisor migration take?
Small estates may move in weeks; large enterprises often plan migrations over many months.
Can we run two hypervisors?
Yes, many organisations do during transition, but long-term dual platforms increase operational cost.
A 90-day evaluation plan
Days 1 to 30: document current virtual machine counts, sizes, guest operating systems, features in use and third-party integrations, and confirm contract renewal dates.
Days 31 to 60: shortlist two or three platforms, build a test cluster for each and migrate representative workloads, including a database and an appliance.
Days 61 to 90: compare results for performance, operations, integration and five-year cost, then present a recommendation with a phased migration plan.
Questions to ask platform vendors
- Which migration tools are provided, and what downtime do they require?
- Which backup, security and monitoring products are certified?
- How are licences or subscriptions priced as clusters grow?
- What is the support model, and where are support engineers located?
- What does the roadmap look like for the next three years?
Key terms explained
- Type 1 platform: virtualisation software that runs directly on server hardware.
- Live migration: moving a running virtual machine between hosts without downtime.
- High availability: automatically restarting virtual machines on another host after a failure.
- Distributed resource scheduling: balancing workloads across hosts automatically.
The bottom line
Licensing changes and new technologies have prompted many organisations to reassess their virtualisation platforms. A disciplined evaluation covers compatibility, operational features, ecosystem support, total cost and vendor health, backed by testing with real workloads. Consider broader alternatives such as public cloud or container platforms for some applications. Whatever the outcome, a phased migration plan with clear rollback options keeps risk under control and protects service levels during the transition.
Further reading on hypervisor
For authoritative, vendor-neutral guidance on hypervisor, see the KVM project. You can also browse our free whitepapers.
