---
title: "Virtualization and Containers"
description: "VMware to Proxmox and Nutanix migrations driven by Broadcom licensing changes, plus Kubernetes platforms built to be operated by your own team."
url: "https://nixit.io/services/virtualization-containers/"
lang: "en"
---
# Virtualization and Containers

We migrate off VMware to Proxmox or Nutanix and build Kubernetes platforms your engineers can actually run.

The change to VMware licensing following the Broadcom acquisition triggered the largest movement in the virtualisation market in a decade. Bundled subscriptions, minimum core count thresholds and the withdrawal of the smaller editions left many small and mid-sized organisations facing renewal quotes several times their previous budget for exactly the same infrastructure. The question is no longer whether alternatives are worth evaluating, but which one fits and what the transition costs.

## Assessment before migration

We do not recommend a platform before examining the environment. That means inventorying virtual machines, comparing allocated resources against actual utilisation, understanding the storage layer, network requirements and every integration with backup, monitoring or orchestration tooling. Those integrations are the most common reason a migration takes longer than the original estimate.

A five-year cost comparison follows, covering licensing, support, hardware and the migration effort itself. Sometimes the conclusion is that renegotiating with VMware remains reasonable, particularly for large estates genuinely using the advanced feature set. For organisations running roughly ten to a hundred hosts, Proxmox or Nutanix almost always come out ahead on the numbers, often by a wide margin.

## Carrying out the migration

Proxmox VE runs on KVM and the Linux kernel, with clustering, automatic failover, live migration and replicated storage via Ceph or ZFS. Nutanix offers a more integrated hyperconverged product with broader commercial support at higher cost, though still below current VMware pricing for many configurations. Which fits depends on your team's Linux depth and on how much you value an integrated management layer.

Migration proceeds in waves. We start with test machines and lower-criticality systems to validate the process and give your team time on the new platform. Virtual machines are converted, drivers and guest agents verified, performance tested under real load, and then cut over. Every wave has a rollback path, and the VMware environment stays operational until the new platform is fully validated.

## Containers and Kubernetes

For applications that genuinely benefit from containerisation, we build Kubernetes platforms designed to be maintained by your engineers rather than only by us. That means authentication integrated with your existing directory, network policies, persistent storage that works under failure, GitOps delivery through Argo CD, and monitoring configured from day one rather than added after the first incident.

We are deliberately cautious about Kubernetes adopted for its own sake. It carries a real operational cost, justified by applications split across many services with frequent releases. For three stable monolithic applications, a well-managed set of virtual machines is simpler, cheaper and considerably easier to diagnose at three in the morning.

## Who this is for

Requests come mainly from organisations that received a VMware renewal quote and are evaluating alternatives, from companies expanding capacity who do not want licensing cost to scale proportionally, and from engineering teams needing a properly operated container platform.

A distinct category is organisations with data sovereignty requirements that prefer infrastructure running on their own hardware within the EU, on open platforms rather than under single-vendor dependency. For these clients the conversation extends beyond cost into long-term control, and being an EU-based partner subject to EU law rather than to third-country access regimes is part of why they engage us.

## Outcomes

Completed migrations typically cut recurring licensing spend substantially while delivering equivalent or better performance, since the underlying hardware is unchanged and the newer hypervisors are efficient. The more durable benefit is structural: your infrastructure is no longer exposed to a unilateral change in one vendor's commercial model, which is precisely the risk that brought most of these clients to us in the first place.

## What you get

- Five-year cost comparison across VMware, Proxmox, Nutanix and other viable options
- Wave-based migration plan with a tested rollback procedure for each phase
- Virtualisation cluster deployed with redundant storage and verified automatic failover
- Kubernetes platform with authentication, networking, persistent storage and GitOps delivery
- Training and operational documentation so your team can run the new platform

## Technologies

Proxmox VE, VMware vSphere, Nutanix AHV, Kubernetes, Ceph, ZFS, Harvester, Argo CD, Longhorn

## Frequently asked questions

### How much did VMware costs actually rise after the Broadcom acquisition?

It varies by customer, but the pattern is consistent: the move to bundled subscriptions with minimum core counts and the removal of smaller editions produced multi-fold increases for small and mid-sized estates. Organisations running a few dozen hosts were hit hardest, and that is the single most common reason clients approach us for migration.

### Is Proxmox mature enough for production?

Yes. It is built on KVM and the Linux kernel, the same foundation most public cloud platforms run on. It provides clustering with automatic failover, live migration, replicated storage through Ceph or ZFS, and commercial support through a subscription. The real differences from VMware are in third-party tooling integrations and in advanced features not every environment uses.

### How long does a VMware migration take and does it require downtime?

For an estate of thirty to fifty virtual machines, four to eight weeks from assessment to completion. Machines move in waves, each with its own window. Most workloads tolerate a short cutover outage; for those that cannot, we use pre-seeded replication that reduces the switch to a few minutes.

### Do we need Kubernetes or are virtual machines enough?

It depends on your applications and your team. Kubernetes earns its operational cost with applications split across many services, frequent deployments and variable scaling needs. For a handful of stable monolithic applications, well-managed virtual machines are simpler and cheaper to run. We do not recommend Kubernetes where it solves no existing problem.