The essentials at a glance

  • Choose based on workload, not platform. Combine the migration with modernization where it makes sense. → Section 2

  • The hypervisor license is only part of the cost. The software in your VMs has its own rules depending on the hypervisor, and those rules can make an inexpensive platform end up being expensive. → Section 7

  • SUE recommends a platform based on an open standard, such as KVM or Kubernetes. Open source is no guarantee against price increases, but it limits the damage. → Section 8

  • Test with a proof of concept. A proof of concept that merely shows a VM booting up proves very little. → Sections 10 and 11

1. Why this is an issue now

General support for vSphere 7 ended in October 2025. For vSphere 8, including vCenter and vSAN 8, general support will end on October 11, 2027 (Broadcom). Anyone still running VMware at that point will be bound by the new licensing terms, which will be more expensive for many organizations. Staying with VMware is also an option—and often the right one for some workloads: with VMware Cloud Foundation 9 or extended support from Broadcom, you’re buying time. Make that decision on a per-workload basis, and before your next renewal.

2. Choose by workload, not by platform

There is no single answer to the question, “What is the best alternative to VMware?” But there is an answer to the question, “Where will this workload run best in the coming years?” Moreover, many workloads are due for modernization. Those who move everything to a new platform all at once often end up paying twice: first for the migration and later for the modernization. Therefore, combine the migration with modernization where it makes sense to do so.

The routes SUE explores in this guide:

Route Examples Right for you if
Other virtualization platform Proxmox VE, Hyper-V, Nutanix AHV, XCP-ng, Apache CloudStack The workload continues to run as a VM, and you mainly want to get rid of the license
Virtualization on Kubernetes SUSE Virtualization, OpenShift Virtualization, KubeVirt You want to manage your VMs and containers on a single platform, or you want to modernize the application later
VMware with a cloud provider Azure VMware Solution, Google Cloud VMware Engine, Amazon Elastic VMware Service You need to move out of your own data center quickly and want to buy some time; the VMware license remains in place
Public cloud AWS, Azure, Google Cloud, European cloud providers Whether the workload has few dependencies or you want to move to the cloud anyway
Containers Kubernetes: Managed or Self-Managed The application is your own software, and you want to continue developing it

Organizations often choose more than one destination. According to a survey by software provider CloudBolt (January 2026, 302 IT decision-makers at North American companies), workloads that are migrated primarily go to the public cloud (72%), to Hyper-V or Azure Stack (38%), and to SaaS (34%) (CloudBolt). As a result, many organizations end up with two or three parallel paths. This isn’t fragmentation, as long as you understand why a workload is located on each path.

3. Decision aid: which situation do you recognize?

Start with your current situation. The table lists, for each situation, the platforms you should look at first, why they’re a good fit, and the biggest risk.

Your situation First, take a look at Why Greatest risk
Specifically, Windows Server, a team with expertise in Microsoft, and hosts that are already licensed for Windows Server Datacenter Hyper-V with Failover Clustering, or Azure Local Hyper-V is included in Windows Server, and the Datacenter edition already supports an unlimited number of Windows VMs per host Linux workloads are receiving less attention; Microsoft's conversion tool will still be in private preview as of November 2026
Linux VMs in particular; licensing costs need to be reduced, and the team is familiar with Linux Proxmox VE or XCP-ng Open source with paid support and a management experience similar to vSphere Microsoft does not fully support Windows on these hypervisors (see “Licensing and Support”)
Kubernetes is already in use or on the roadmap, and applications are being modernized OpenShift Virtualization, SUSE Virtualization VMs and containers on a single platform, with the same tools, RBAC, and GitOps Your team manages Kubernetes, including for workloads that are never containerized
A private cloud with self-service capabilities for multiple teams, departments, or customers Apache CloudStack Multitenancy, quotas, self-service, and APIs are at the core of the product More design work upfront and a smaller market for expertise and support
You want a single provider for hypervisors, storage, microsegmentation, and management Nutanix AHV A complete package, with a management experience similar to vCenter You're just trading one dependency for another
Your contract will expire within a year, or your data center will have to close Renewing VMware with a Cloud Provider Save time without having to rebuild VMs You'll be paying for VMware for longer; without an exit strategy, this interim step will become permanent
A European supplier or open governance is a requirement Proxmox VE, XCP-ng, SUSE Virtualization, or Apache CloudStack Suppliers headquartered in the EU, or a project under the Apache Software Foundation Smaller suppliers: Check their support and partner network in the Netherlands

Do you recognize more than one situation? That's normal. In that case, split up your workloads and choose a path for each group.

4. The platforms side by side

This comparison reflects the situation as of November 2026. Vendors often update their products, licenses, and support terms, especially following an acquisition. Be sure to check the details with the vendor before making a decision.

Platform Underlying Technology Licensing model Supplier support VMs and containers on a single platform Migration from VMware Backup (Veeam, Commvault)
Proxmox VE KVM on Debian Open source; paid support per CPU socket Yes, Proxmox (Vienna) VMs and LXC containers, not Kubernetes Built-in import wizard for ESXi 6.5 through 8.0 Both
SUSE Virtualization (formerly Harvester) KubeVirt, KVM Open source; subscription for support Yes, SUSE Yes, manage via Rancher Built-in VM Import Controller Veeam Cabinets (supported by Veeam, documented by SUSE)
OpenShift Virtualization KubeVirt, KVM Subscription; Virtualization Engine per node, for VMs only Yes, Red Hat Yes, with a full OpenShift subscription Migration Toolkit for Virtualization Veeam (version 13.1 and later)
KubeVirt KVM on Kubernetes Open source (CNCF) No, only through distributions Yes No proprietary tool Depending on the distribution
Nutanix AHV AHV Per core Yes, Nutanix Via a separate Kubernetes platform Nutanix Move Both
Hyper-V (Windows Server 2025) Hyper-V Per core, at least 16 per server Yes, Microsoft Windows containers VM Conversion in Windows Admin Center (private preview in November 2026) Both
Azure Local Hyper-V, managed through Azure Per physical core, as a subscription Yes, Microsoft Yes, with AKS Azure Migrate Veeam, Commvault
Apache CloudStack Usually KVM Open source (Apache) Through specialized providers, such as ShapeBlue No Improved in version 4.22 Check with your backup provider
KVM with libvirt KVM on Linux (RHEL, Ubuntu, SUSE) Open source; support provided through a subscription to the Linux distribution Through Red Hat, Canonical, or SUSE No virt-v2v Check with your backup provider
XCP-ng Xen Open source; support via Vates Yes, Vates No Built into Xen Orchestra Veeam

Sources: Proxmox, SUSE, Red Hat, Nutanix, Microsoft, Azure Local, CloudStack, Veeam, Commvault.

5. Specific VMware features and their equivalents

The first question engineers ask is: What replaces vMotion, DRS, vSAN, and NSX? This table lists the closest equivalent for each platform. An equivalent isn’t always identical; test the features you actually use in the proof of concept.

VMware Proxmox VE XCP-ng Hyper-V Nutanix AHV OpenShift Virtualization SUSE Virtualization Apache CloudStack
vMotion Live Migration Live migration, including storage Live Migration Live Migration Live Migration Live Migration Live Migration (KVM)
HA HA Manager with Fencing HA in Xen Orchestra Failover Clustering Built-in Kubernetes restarts VMs on a different node Kubernetes restarts VMs on a different node Built-in
DRS Dynamic load balancer (since 9.2) Load Balancer Plugin in Xen Orchestra Dynamic Optimization in System Center VMM Acropolis Dynamic Scheduling Kubernetes Scheduler and Descheduler Kubernetes Scheduler DRS by cluster
vSAN Ceph (integrated) or ZFS with replication XOSTOR Storage Spaces Direct Built-in distributed storage OpenShift Data Foundation or CSI Storage Longhorn or external CSI storage External primary storage, such as Ceph, NFS, or iSCSI
NSX SDN with VXLAN and EVPN, firewall per VM SDN Controller in Xen Orchestra SDN with Network Controller and Data Center Firewall Flow Network Security (microsegmentation) OVN-Kubernetes with network policies and additional networks VLAN Networks; Network Policies via Kubernetes Advanced zones with VPCs and security groups
vCenter Web interface per cluster, Proxmox Datacenter Manager for multiple clusters Xen Orchestra Windows Admin Center or System Center VMM Prism Central OpenShift Console Harvester UI and Rancher CloudStack UI with self-service
PowerCLI and Aria REST API, Terraform (community), and Ansible API, Terraform Provider, and Ansible PowerShell, and Azure tools for Azure Local as well API, Terraform Provider, and Ansible Kubernetes API, GitOps, and Ansible Kubernetes API, Terraform Provider API, Terraform Provider, and Ansible

This table shows the status as of November 2026, based on documentation provided by the vendors. The biggest differences are in the network. Nutanix Flow comes closest to microsegmentation with dynamic groups, as found in the NSX distributed firewall. On other platforms, this is often implemented using network policies, security groups, or a separate firewall.

6. By platform: when is it appropriate?

Proxmox VE

Proxmox VE runs VMs on KVM and lightweight containers on LXC, with its own web interface. The software is open source; paid support is billed per CPU socket, not per core. Since version 9.0, snapshots have also worked on shared SAN storage (as a technology preview), and Proxmox Datacenter Manager (December 2025) manages multiple clusters from a single screen.

  • Ideal for: organizations that primarily run VMs and want to reduce licensing costs, with a team that is familiar with Linux.
  • Please note: Proxmox does not have a support agreement with Microsoft, so Microsoft does not fully support Windows VMs on Proxmox. The firewall works on a per-VM basis using security groups; microsegmentation, such as that found in NSX, must be configured differently. Dynamic load balancing between hosts has only been built-in since version 9.2.

SUSE Virtualization

SUSE Virtualization (formerly Harvester) runs VMs on Kubernetes using KubeVirt and manages them alongside containers via Rancher. In addition to its own storage (Longhorn), you can use external storage from other vendors via CSI drivers.

  • Ideal for: organizations that want to run VMs and containers on a single platform, or that already use Rancher.
  • Please note: Your team must manage Kubernetes. SUSE only lists Veeam Kasten as a backup partner.

Red Hat OpenShift Virtualization

OpenShift Virtualization runs VMs as part of OpenShift. The Virtualization Engine edition is for VMs only and is licensed per node; for containers, you need a full OpenShift subscription. The Migration Toolkit for Virtualization is included and, starting with version 2.11, delegates data transfer during a cold migration to supported storage arrays (for a warm migration, this is still a technology preview).

  • Suitable for: organizations that are already running OpenShift.
  • Please note: Virtualization Engine only supports VMs; if you also want to run containers, you’ll need a different edition. Veeam Data Platform 13.1 (July 2026) adds support for OpenShift Virtualization; the corresponding plug-in became available later, so check the status for your version.

KubeVirt

KubeVirt is the open-source project behind SUSE Virtualization and OpenShift Virtualization. You can also use it independently on any Kubernetes cluster.

  • Best suited for: teams with extensive Kubernetes experience who build their own platforms.
  • Please note: There is no vendor support, and you are responsible for arranging migration tools and backups yourself.

Talos Linux (new, not yet ready for production)

Sidero Labs is building a hypervisor into Talos Linux, the minimal operating system for Kubernetes. VMs will then run without Kubernetes and without KubeVirt, using the same declarative management as the rest of Talos. As of October 2026, the hypervisor was in alpha; the final version is scheduled for December 2026. The commercial management platform Talos Director, featuring live migration, high availability (HA), and migration from VMware, will follow in January 2027. Sidero Labs has been owned by the U.S.-based company Yardi since September 2026.

  • Ideal for: teams that are already using Talos for Kubernetes and want to run VMs under the same management. This is already possible today with KubeVirt on a Talos cluster.
  • Please note: No information is currently available regarding Microsoft support for Windows VMs, backup, or certification by software vendors. Wait until the final version is released before testing.

Nutanix AHV

Nutanix provides its own hypervisor, managed through Prism, which is licensed per core. AHV now also supports external storage from vendors such as Dell and Pure Storage. Nutanix Move migrates VMs from ESXi and replaces VMware Tools.

  • Ideal for: organizations that want a complete platform from a single vendor, with a management experience similar to vCenter.
  • Please note: You're simply switching from one provider to another. Compare the total costs over five years, not just the first year.

Hyper-V and Azure Local

Hyper-V is included in Windows Server 2025. With the Datacenter edition, you can run an unlimited number of VMs on a single server. Azure Local (formerly Azure Stack HCI) is the version you run on-premises and manage through Azure; it is licensed per physical core as a subscription.

  • Best suited for: organizations with a large number of Windows workloads and a Microsoft-focused team.
  • Please note: The tool for converting VMware VMs has been available in Windows Admin Center’s vMode since November 2026, but is still in private preview; you can sign up through Microsoft’s Technology Adoption Program. For Azure Local, there is Azure Migrate. When migrating to Azure Local, static IP addresses are retained only for Windows VMs.

Apache CloudStack

CloudStack is an open-source cloud platform for multitenant environments, typically using KVM. It is designed to build a private cloud or a service for multiple customers.

  • Suitable for: large private clouds and environments with multiple departments or customers.
  • Please note: CloudStack requires more design work than a standalone hypervisor.

KVM with libvirt

KVM with libvirt forms the foundation of Proxmox VE, OpenShift Virtualization, and SUSE Virtualization, without a management layer on top. You manage hosts and VMs using libvirt, Cockpit, or virt-manager, and automate tasks with Ansible or Terraform. You can migrate from VMware using virt-v2v.

  • Best suited for: teams with extensive Linux knowledge who want to manage a manageable number of VMs as code.
  • Please note: There is no central management environment like vCenter; you must configure HA, live migration, and backup yourself.

XCP-ng

XCP-ng is an open-source hypervisor based on Xen, developed by the French company Vates, with Xen Orchestra as its management environment. Veeam has supported it since version 13.1.

  • Suitable for: organizations looking for a European provider of traditional virtualization.
  • Please note: The ecosystem is smaller than that of KVM platforms, and Vates does not have a support agreement with Microsoft for Windows VMs.

VMware with a cloud provider

With Azure VMware Solution, Google Cloud VMware Engine, or Amazon Elastic VMware Service, you can run VMware in the cloud without having to convert your VMs. As of November 1, 2025, you’ll need to bring your own VMware Cloud Foundation license, which you purchase from Broadcom (Google Cloud).

  • Ideal for: organizations that need to move out of their data center quickly and want to buy time.
  • Please note: You will still be dependent on the VMware license, so this is an interim step.

Public cloud or containers

The public cloud is the most commonly cited destination: 72% in the CloudBolt survey (CloudBolt). Tools such as AWS Transform for VMware help with inventory and translating network configurations.

  • Suitable for: workloads with few dependencies, and applications that you still want to modernize.
  • Please note: Moving a VM to the cloud without making any changes is rarely cheaper. Read the guide at Cloud Migration for more information.

7. Licenses and support not included in the quote

The hypervisor license is only part of the cost. The software in your VMs has its own rules depending on the hypervisor, and those rules can make an inexpensive platform end up being expensive.

Windows Server: your licenses come with you

Microsoft licenses Windows Server per physical core of the host, regardless of the hypervisor. With the Datacenter edition, you can run an unlimited number of Windows VMs on a fully licensed host (Microsoft). If you’ve already licensed your VMware hosts with Datacenter, those licenses remain valid on the same hardware, regardless of which platform you choose. Hyper-V therefore does not require an additional license on those hosts. Please note: the free Hyper-V Server is no longer available; the last version was released in 2019.

Microsoft support for Windows VMs

Microsoft fully supports its software on hypervisors from vendors that have a support agreement for virtualization. That list includes Canonical, Citrix, Nutanix, Red Hat, SUSE, Virtuozzo, and VMware. Proxmox and Vates (XCP-ng) are not on the list (Microsoft).

On a hypervisor not included in that list, Microsoft will only investigate issues for customers with a Premier support contract. Microsoft may then require you to first reproduce the issue on Hyper-V or on physical hardware (Microsoft). For a web server, this is manageable. For SQL Server or a business-critical Windows cluster, you consciously accept this risk and document it in your decision-making document.

Oracle Database: license the entire cluster

Oracle considers VMware to be soft partitioning. This means that you license all physical cores on which an Oracle VM can run, not just the VM’s vCPUs. Of the KVM variants, Oracle only recognizes Oracle Linux KVM as hard partitioning, and then only with statically assigned cores (Oracle). Proxmox, OpenShift, SUSE Virtualization, and Hyper-V are not on Oracle’s list. Therefore, deploy Oracle Database on its own small cluster, or account for the cores of the entire cluster on which the VM can run.

SAP and other software vendors

SAP certifies on a per-hypervisor basis, and other software vendors often have their own lists. Request a support statement for each business-critical application before you choose a platform, not after.

8. Supplier, ownership, and sovereignty

Broadcom's acquisition of VMware showed that the owner of your platform determines your costs. That's why you should also consider who is behind a platform and what happens if that owner changes course.

Platform Supplier Headquarters Ownership or governance Open source code
Proxmox VE Proxmox Server Solutions Vienna, Austria Private Yes (AGPLv3)
XCP-ng Vates Grenoble, France Private Yes
SUSE Virtualization SUSE Nuremberg, Germany Investor EQT Yes
OpenShift Virtualization Red Hat United States IBM Yes (KubeVirt)
KubeVirt Cloud Native Computing Foundation N/A Open Governance Yes
Apache CloudStack Apache Software Foundation N/A Open Governance Yes
Hyper-V and Azure Local Microsoft United States Publicly Traded No
Nutanix AHV Nutanix United States Publicly Traded No

Open source is no guarantee against price increases, but it limits the damage: if the vendor changes its terms, you can continue to run the software or purchase support from another provider. With proprietary software, you don’t have that option. That’s why SUE recommends a platform based on an open standard, such as KVM or Kubernetes. If you do choose Hyper-V, Azure Local, or Nutanix, do so deliberately: you’re trading one dependency for another, and any future switch will require another migration.

9. What often proves disappointing in migrations

In SUE migrations, most of the work isn't in the hypervisor, but in backup, monitoring, and the scripts related to vCenter. The same points keep coming up in analyses by Gartner, on the Proxmox forum, and in published migration reports.

Networking and NSX. According to Gartner, migrating networks and their security is more challenging than migrating the hypervisor (The Register). Anyone using microsegmentation, distributed firewalls, or overlay networks in NSX will need to redesign them. Take stock of your NSX rules before choosing a platform.

Drivers and startup issues. On the Proxmox forum, “inaccessible boot device” is a recurring error message: Windows VMs fail to boot after migration because the VirtIO drivers are missing or the storage controller has changed (Proxmox forum). The transition from BIOS to UEFI also requires attention. Install drivers in advance and test each type of VM.

Backup and disaster recovery. You’ll need to reconfigure backup jobs, replication, and recovery tests, even if you stick with the same backup provider. Check which backup solution each platform officially supports.

Storage to replace vSAN. Hyperconverged storage with triple replication leaves about one-third of the raw capacity available. At the Portuguese company Centrogest, out of 72 TB of raw NVMe capacity in a three-node Proxmox cluster, approximately 24 TB remained available for VMs (dev.to). Reusing an existing SAN may therefore be more cost-effective.

Monitoring, automation, and runbooks. Scripts in PowerCLI, workflows in Aria, and monitoring in vCenter no longer work. This work is often underestimated during planning.

HA and load balancing. High availability, DRS, and live migration work differently on each platform. Test failover scenarios before migrating the first production workloads.

Support from software vendors. Not every software vendor supports every platform. For example, on the Proxmox forum, administrators report that SAP Business One is not supported on Proxmox (Proxmox forum). Request support statements for your business-critical applications.

Team expertise. Each alternative requires different expertise than vSphere: Linux and KVM, Kubernetes, or Windows Server and Azure. Plan for training or outside assistance from the start, not only after the first migration batch gets stuck.

Turnaround time. Gartner estimates 18 to 48 months for organizations with more than 2,000 VMs and at least 100 servers (The Register). How quickly the batches are processed often depends on the bandwidth between the old and new environments.

10. How to choose in five steps

 

  1. Map out your environment. Count VMs, cores, storage, and network dependencies, and note which VMware features you actually use: NSX, vSAN, DRS, Aria.
  2. Allocate workloads by route. For each group, determine whether it will remain as is, move to another on-premises platform, move to the public cloud, or be modernized.
  3. Check the mandatory requirements for each platform: support from your software vendors, support from your backup solution, storage integration, and your team’s expertise.
  4. Calculate the costs over a five-year period. Include licenses, support, hardware, migration, training, and management—not just the license price.
  5. Conduct a proof-of-concept test. Migrate a representative set of VMs, including a challenging Windows VM and an application with network dependencies, and test failover and recovery. The test plan below outlines the scenarios.

Do you have little time left before your current VMware contract expires? If so, consider an interim solution—such as a shorter renewal or VMware through a cloud provider—so you can plan the migration without rushing.

11. Test plan for the proof of concept

A proof of concept that merely demonstrates that a VM boots up doesn't prove much. Test the scenarios in which your current environment protects you, and measure the results so you can compare platforms fairly.

Test What you do What you measure
Migrating the Most Challenging VMs Migrate the largest VM, the oldest Windows VM, and a VM with network dependencies Downtime during the cutover, manual work after the migration
Host is down Cut the power to a host—don't shut it down properly It's been a while since the VMs were running again—and will they ever come back?
Live Migration Under Load Move a VM while the application is under load Interruption for users or in the logs
Storage is down Remove a disk or node from a hyperconverged storage system Rebuild time and its impact on latency during the rebuild
Restore from a backup Restore an entire VM and individual files Recovery time relative to your RTO
Cluster Upgrade Perform a rolling upgrade Whether VMs continue to run
Integration Integrate AD or LDAP, RBAC, and monitoring, and create a VM using Terraform or Ansible What Works Without Customization
Performance Run fio or an application benchmark on the old and new platforms Latency and Throughput with the Same Hardware
Support Open an official ticket with the vendor or partner Response Time and Quality of the Answer

Preferably, use the same hardware for each platform you test so that any differences do not stem from the hardware. Define in advance what constitutes a successful test.

12. Here's how to present the decision internally

A CTO makes decisions based on risk, cost, and feasibility. A two-page decision document covering these elements usually works better than a detailed report:

  1. Background. The expiration date of your VMware contract and support, and what happens if you don't take any action.
  2. Scope. Number of VMs, hosts, and cores, and the organization into workload groups with their routes.
  3. Options. Two or three platforms that you've thoroughly tested, plus the option to extend as a reference.
  4. Evaluation. A weighting matrix showing the results of the proof of concept.
  5. Costs over five years. Per option, including all items listed below.
  6. Risks. What can go wrong and what steps you should take, including support from Microsoft, Oracle, and SAP.
  7. Planning and Decision. Phasing: when you’ll be free of VMware and what decision you’re asking for now.

Example of a weighting matrix

The weighting below is a starting point. Adjust it to fit your situation: an organization with a lot of Oracle systems will place greater emphasis on supportability, while an organization with a small team will place greater emphasis on expertise.

Criterion Question you're answering Example weighting
Scalability of Your Applications Do Microsoft, Oracle, SAP, and your other software vendors support this platform? 25%
Features You Actually Use Do HA, networking, storage, and live migration work the way you need them to? 20%
Expertise within the team Can your team handle this, and can you find people with this expertise? 15%
Costs over five years What is the total cost, including migration and parallel operation? 15%
Migration Risk and Lead Time How quickly can you migrate away from VMware, and how much downtime will it involve? 10%
Supplier How high is the risk of lock-in or a change in exchange rate? 10%
Backup, Monitoring, and Automation Do your current tools work, or do you need to replace them as well? 5%

Cost items that are often omitted

 

  • Licenses or subscriptions, and support from the vendor or partner.
  • Hardware: Can you reuse existing servers? Are they on the compatibility list? And do you need additional storage capacity for replication?
  • Migration: tooling, hours, and testing per application.
  • Double payment period: the time during which you pay for both VMware and the new platform at the same time.
  • Replacement or modification of backup, monitoring, and automation systems.
  • Training and management capacity after the migration.
  • Software licenses, such as those for Oracle or Windows, which may change depending on the choice made.

13. What SUE does

SUE supports workload migrations to platforms such as Red Hat OpenShift Virtualization, SUSE Virtualization, Apache CloudStack, Proxmox VE, and KVM with libvirt, and implements those platforms. In addition, we support migrations to the public cloud (AWS, Microsoft Azure, and Google Cloud) and to containers. Nutanix, Hyper-V, and Azure Local are also options. The approach we recommend depends on your workloads, your team, and your requirements.

We also help with Application Modernization, where we can first map out the existing code using Re:App. We can then manage the new platform using Managed Services for Linux and Kubernetes. That way, you don’t pay twice—first for the migration and later for the modernization.

For example, NXP replaced its proprietary virtualization software with Apache CloudStack. In a two-month proof of concept, CloudStack met 95% of NXP’s more than 40 functional requirements. Afterward, SUE helped the NXP team replicate the deployment across its global environments; NXP’s new data center in the U.S. runs entirely on Apache CloudStack (customer case NXP). In April 2026, SUE organized a TechTalk on OpenShift Virtualization, and SUE students wrote research papers on the transition from VMs to containers. We are currently helping several clients transition to Proxmox, OpenShift, SUSE, the cloud, and containers.

With our free Cloud Readiness Assessment, we’ll map out your VMware environment within two weeks and determine the best path forward for each workload.

14. Frequently asked questions

All frequently asked questions
There is no single best alternative for every organization. According to SUE, exiting VMware is a portfolio decision: stay with each workload as-is, migrate it as a VM, or modernize it, and then run it on an open standard that you can leave at any time. Proxmox VE and Hyper-V are suitable for traditional virtualization; SUSE Virtualization and OpenShift Virtualization are suitable for organizations that want to combine VMs and containers; and Apache CloudStack is suitable for large private clouds. Choose a solution for each group of workloads based on support, backup, storage, and your team’s expertise.
General support for vSphere 8 ends on October 11, 2027. Support for vSphere 7 ended in October 2025. After that, you will no longer receive security updates unless you sign an extended support contract with Broadcom.
That depends on the size and complexity. Gartner estimates 18 to 48 months for environments with more than 2,000 VMs. Smaller environments can be set up more quickly, but network, backup, and testing take the most time regardless of size.
Yes. Proxmox VE is running in production with paid support from Proxmox itself, and backup can be handled using Veeam and Commvault, among other solutions. Be aware of features that are missing compared to VMware, such as NSX microsegmentation, and check whether your software vendors support Proxmox.
Yes. SUSE Virtualization and OpenShift Virtualization run VMs on Kubernetes using KubeVirt. This allows you to first migrate applications as VMs and then gradually transition them to containers, all on the same platform.
As an interim step, yes. With Azure VMware Solution, Google Cloud VMware Engine, or Amazon Elastic VMware Service, you can quickly move out of your data center without having to migrate your VMs. You will, however, remain dependent on a VMware license, which you’ll purchase directly from Broadcom starting in November 2025.
Yes. Microsoft licenses Windows Server per physical core of the host, regardless of the hypervisor. If you continue to use the same hardware, your licenses remain valid. With the Datacenter edition, you can run an unlimited number of Windows VMs per licensed host, including on Proxmox, Nutanix, or OpenShift.
Not entirely. Microsoft supports Windows on hypervisors from vendors with a support agreement, such as Red Hat, SUSE, and Nutanix. Proxmox and Vates (XCP-ng) are not included in this list. Microsoft investigates issues only for customers with Premier Support and may ask them to reproduce the issue on Hyper-V first.
Oracle only recognizes Oracle Linux KVM as hard partitioning. On VMware and other KVM platforms, you must license all physical cores on which an Oracle VM can run. Therefore, deploy Oracle Database on its own small cluster so that you do not have to license the cores of your entire new platform.

Schedule a free Cloud Readiness Assessment

With our free Cloud Readiness Assessment, we’ll map out your VMware environment within two weeks and determine the best path forward for each workload. Schedule an assessment

Stefan Behlen
Stefan Behlen

Let’s talk


Schedule a free Cloud Readiness Assessment

With our free Cloud Readiness Assessment, we’ll map out your VMware environment within two weeks and determine the best path forward for each workload. Schedule an assessment

* required

We will respond within one business day. By submitting this form, you confirm that you have read and understood our privacy policy.
SUE Cloud & AI Native Solutions
Privacy overview
This website uses cookies. We use cookies to ensure that our website and services function properly, to gain insight into the use of our website, and to improve our products and marketing. For more information, please read our privacy and cookie policy.