Openshift Virtualization: Features, Architecture, and Tutorial

What Is Red Hat OpenShift Virtualization? 

OpenShift Virtualization is a Red Hat OpenShift feature that lets you run traditional Virtual Machines (VMs) alongside containers and serverless functions on a single platform, using KVM and KubeVirt to manage them like native Kubernetes objects for a unified hybrid cloud experience, simplifying migration and modernization for mixed workloads. It bridges the gap between legacy VM environments and modern cloud-native applications, allowing for integrated management, networking, and storage.

Key features and benefits:

• Unified platform: Manage VMs and containers on the same OpenShift cluster, reducing operational complexity.
• Container-native approach: VMs run as pods, leveraging Kubernetes’ orchestration for lifecycle management (start, stop, restart).
• Hybrid cloud mobility: Easily migrate and run VMs across on-premises and public cloud environments.
• Modernization: Transition legacy VM applications to a cloud-native platform without a complete rewrite.
• Integrated ecosystem: Use familiar OpenShift tools for compute, storage, and networking for both workloads.
• Supports diverse architectures: Now supports ARM-based hardware for unified management across different CPU types.

Common use cases:

• Bridging gaps: Running older VM applications that need to interact with newer containerized services.
• Disaster recovery: Implementing robust backup, replication, and failover for VMs.
• Infrastructure consolidation: Running diverse workloads (VMs, containers, AI) on shared hardware for efficiency.

This is part of a series of articles about Red Hat OpenShift

Key Features and Benefits of OpenShift Virtualization 

Unified Platform

OpenShift virtualization delivers a single control plane for VMs and containers, simplifying day-to-day management across both workload types. Operators can provision, deploy, and scale containerized applications and virtual machines using the same OpenShift web console and command-line tools. 

Operational consistency is further supported by unified networking, storage, and security policies. Teams gain granular control over resources and permissions, simplifying compliance and governance processes. End-to-end automation and observability across all workloads enhance developer productivity and simplify troubleshooting.

Container-Native Approach

OpenShift virtualization leverages Kubernetes primitives and APIs to manage virtual machines as first-class citizens. Each VM becomes a Kubernetes object, benefiting from features like namespaces, RBAC, and resource quotas. This approach allows teams to manage VMs through the same CI/CD pipelines, GitOps workflows, and automation strategies used for containers, maximizing agility and accelerating software delivery.

Native integration with container technologies enables hybrid workloads, such as coupling legacy applications inside VMs with modern microservices. Developers can use Kubernetes-native tools like Operators and Helm charts to manage, update, and observe both virtual machines and containers.

Hybrid Cloud Mobility

With OpenShift virtualization, workloads can be moved fluidly between on-premises and cloud environments. VMs and containers coexist on the same clusters and can migrate between different infrastructures without the complexity of traditional virtualization stacks. This mobility helps organizations avoid vendor lock-in, optimize costs, and respond rapidly to changing business needs or compliance requirements.

Live migration capabilities further support workload mobility, reducing downtime during infrastructure maintenance or scaling events. Workloads can be balanced across clusters or moved to public clouds for disaster recovery, seasonal scaling, or geographic optimization. 

Modernization

OpenShift virtualization enables gradual modernization of legacy workloads by letting teams run existing VMs on a Kubernetes-native platform. Applications that are not yet container-ready can continue to operate in VMs, while new workloads or microservices are natively deployed in containers, all managed side-by-side. 

As organizations rewrite or repackage legacy applications, they can decommission VMs and shift to containers seamlessly on the same platform. The unified control plane accelerates modernization initiatives and bridges gaps between infrastructure teams, allowing for faster delivery cycles, reduced technical debt, and lower operating costs compared to managing separate environments.

Integrated Ecosystem

OpenShift virtualization benefits from tight integration with Red Hat’s certified ecosystem of tools and partner technologies. Networking, storage, monitoring, and security solutions from Red Hat and third parties are available out-of-the-box. Existing OpenShift clusters can leverage their ecosystem investments to support VM workloads, extending mature DevOps and monitoring workflows directly to legacy applications.

This integrated environment simplifies vendor management and reduces friction when scaling or optimizing workloads. Teams can use familiar OpenShift management interfaces, benefit from Red Hat support, and capitalize on tested integrations with backup, compliance, and automation tools, ensuring production readiness for both containers and VMs.

Supports Diverse Architectures

OpenShift virtualization is designed for flexibility across multiple deployment scenarios, including private, public, and hybrid clouds. It can run on physical or virtual infrastructure and supports a range of hardware architectures and acceleration features through integration with technologies like SR-IOV for high-performance networking or GPUs for specialized workloads.

Support for diverse architectures ensures that organizations can optimize application performance, leverage existing hardware investments, and accommodate specialized workloads such as AI/ML, NFV, or high-throughput transactional applications. This flexibility supports growth and innovation without locking organizations into a single infrastructure model.

Related content: Read our guide to OpenShift pricing (coming soon)

How OpenShift Virtualization Architecture Works 

KubeVirt Control Plane

The core of OpenShift virtualization is built on KubeVirt, which runs several custom Kubernetes controllers and components to manage virtual machines. The primary control plane components include virt-controller, which manages the lifecycle and orchestration of VM resources, and virt-handler, which runs on every node in the cluster to monitor and support active VMs. These components ensure VMs are instantiated, monitored, and cleaned up according to Kubernetes lifecycle events, leveraging established container orchestration primitives.

virt-launcher is responsible for launching and managing VM processes isolated inside Kubernetes pods. On the node, libvirtd and qemu-kvm provide the necessary virtualization backend, enabling hardware acceleration and direct interaction with VM operating systems. 

Custom Resources

OpenShift virtualization extends Kubernetes’ API with several custom resources for managing virtual machines and their associated assets. The VirtualMachine resource describes the configuration and intended state of a VM, while the VirtualMachineInstance defines the running state and lifecycle of the VM on the cluster. These custom resources expose familiar declarative management, enabling VMs to be integrated into GitOps workflows and infrastructure-as-code pipelines.

Other custom resources like DataVolume enable dynamic creation and population of persistent storage for VMs, supporting workflows such as image imports, cloning, and backups. Instancetype and Preference CRs provide VM sizing templates and user preferences, standardizing deployments and ensuring consistent configurations across teams or projects. 

Networking Model

Networking for OpenShift virtualization leverages both Kubernetes-native and advanced networking features. Multus, a CNI plugin, enables attaching multiple network interfaces to pods running VMs, allowing for separate control and data plane networks or integration with legacy VLANs. Pods use the default pod network for management traffic, while additional networks can be set up using custom bridges, bonds, or direct SR-IOV attachment for high-performance requirements.

Specialized networking options, including live-migration networks, enable seamless VM migrations with minimized downtime and data loss. These configurations are critical for enterprise workloads that require high availability, guaranteed bandwidth, or complex topologies. Integrating these capabilities within the OpenShift networking stack ensures that both VMs and containers can share and manage networking resources efficiently.

Storage Model

OpenShift virtualization provides storage integration through Kubernetes persistent volume claims (PVCs), allowing VMs to attach persistent block or file storage managed by OpenShift. The containerized data importer (CDI) service simplifies importing VM images, cloning PVCs, and automating disk provisioning workflows. VolumeSnapshots enable point-in-time backups and rapid recovery of VM disks, essential for data protection and disaster recovery scenarios.

Storage access modes, like ReadWriteOnce (RWO) and ReadWriteMany (RWX), dictate whether a PVC can be mounted by a single VM or shared across multiple instances or pods. Boot-from-PVC enables VMs to start directly from persistent volumes, supporting consistent and repeatable deployments. These storage models ensure robust lifecycle management, efficient data migration, and reliability for both legacy VMs and containerized applications.

Scheduling and Placement

OpenShift virtualization takes advantage of Kubernetes advanced scheduling features to optimize VM placement and performance. NUMA (non-uniform memory access) awareness ensures optimal memory access and compute efficiency for workloads sensitive to hardware topology. The CPU Manager and Topology Manager work together to assign dedicated CPUs and align memory allocations, catering to performance-critical or latency-sensitive applications.

Administrators can use node labels, taints, and tolerations to fine-tune placement of VMs, assigning workloads only to appropriately configured nodes or isolating specialized workloads. These tools allow teams to segment infrastructure, enforce compliance, and optimize resource utilization, ensuring that both containers and VMs can run efficiently and reliably at scale.