In today's digital-first era, businesses are driven by massive amounts of data. As data volumes scale and environments modernize toward cloud-native architectures, safeguarding mission-critical workloads while maintaining steady operations becomes vital. To meet defined recovery time objective (RTO) and recovery point objective (RPO) targets, organizations need a robust high availability and disaster recovery strategy that enables trustworthy, rapid recovery while preventing or minimizing data loss.

This white paper presents the unified high availability and disaster recovery solution for safeguarding VMs running on the Red Hat OpenShift Virtualization platform with Dell PowerFlex software-defined infrastructure. The solution provides storage-native, policy-driven replication and resiliency by using the PowerFlex Container Storage Interface (CSI) driver along with the Dell Container Storage Modules (CSM) Replication and Resiliency, enabling enterprise class high availability (HA) and disaster recovery (DR) capabilities.

Dell PowerFlex delivers high-throughput, linearly scalable, fault-tolerant primary storage through protection domains and fault sets. Red Hat OpenShift uses Policy-as-Code (PaC) to enforce protection through native Kubernetes construct to provide consistent security and auditable operations across hybrid cloud environments. Together, these technologies deliver a unified, cloud-native protection fabric that supports both local resilience and multisite recovery to safeguard critical data, streamline runbooks with GitOps-ready workflows, and sustain business continuity.

This white paper is intended for technical professionals and decision makers responsible for the design, deployment, and operation of enterprise virtualization and high availability (HA) and disaster recovery (DR) solutions. The primary audience includes database administrators, system engineers, field consultants, IT administrators, technical architects, and infrastructure leaders involved in implementing HA and DR for VMs running on OpenShift Virtualization using Dell PowerFlex CSM Replication and Resiliency modules.

Readers are expected to possess foundational knowledge of PowerFlex, Red Hat OpenShift Container Platform (OCP), OpenShift Virtualization, and modern container technologies. Familiarity with core infrastructure domains such as storage, compute, and networking architectures is assumed.

In addition, readers should have a working understanding of enterprise data protection concepts such as storage replication, fault tolerance, high availability, backup and recovery, and disaster recovery. This context is essential to fully evaluate the architectural considerations, design decisions, and operational implications discussed throughout this white paper.

Dell Technologies and the authors of this document welcome your feedback on the solution and the solution documentation. Contact the Dell Technologies Solutions team by email.

Author

Vinutha Shenoy

PowerFlex software-defined storage enables broad consolidation across the data center, encompassing most block storage workloads with ease. The software-first architecture enables automation and programmability of the complete infrastructure stack. PowerFlex provides scalability, performance, and resiliency to enable effortless adherence to stringent workload SLAs.

The PowerFlex family provides a foundation that combines compute and high-performance storage resources in a managed, unified fabric. PowerFlex comes in flexible consumption options (rack, appliance, and custom nodes) that enable various independent, disaggregated architectures. PowerFlex is ideal for high-performance applications and databases, building an agile private/hybrid cloud, or consolidating resources across heterogeneous environments.

Software is the key differentiation in the PowerFlex offer. PowerFlex software components not only provide the distributed software-defined storage services but also help simplify infrastructure management and orchestration.

PowerFlex is a software-defined scale-out block storage service that is designed to deliver flexibility, elasticity, and simplicity with predictable high performance and resiliency at scale.

For more details about PowerFlex 4.x, see PowerFlex 4.x specifications.

For more details about PowerFlex 5.x, see PowerFlex 5.x specifications.

PowerFlex Manager enables comprehensive IT Operational Management (ITOM) and Life Cycle Management (LCM) capabilities that span compute as well as storage infrastructure, from BIOS and firmware to nodes, software, and networking.

PowerFlex software-defined storage can be deployed in a disaggregated architecture (independently scalable compute and storage layers).

In an independent, or disaggregated, architecture some nodes provide storage capacity for data in applications, while other separate and independent nodes provide compute resources for applications and workloads. Compute and storage resources can be scaled independently by adding nodes to the cluster while it remains active. This separation of compute and storage resources helps to minimize software licensing costs in certain situations. This architecture is ideal for high-performance databases and application workloads.

Disaggregation gives organizations the freedom to choose different LCM options that best align with their operational models. Not only can PowerFlex Manager manage the LCM of PowerFlex storage infrastructure, it can also manage LCM of PowerFlex compute nodes. With this freedom, there is also the choice to manage compute estates that connect to PowerFlex with other tooling.

PowerFlex rack is a software-defined infrastructure platform that combines compute and high-performance storage resources in a managed unified network. This rack-based engineered system, with integrated networking, enables customers to achieve the scalability and management requirements of a modern data center.

PowerFlex appliance is a PowerEdge-based server which has been configured to be a node in a software-defined storage deployment that runs PowerFlex software components. This offering allows customers the flexibility and savings to bring their own compatible networking or make use of full network automation using a supported set of Dell or Cisco switches.

PowerFlex Custom Nodes are validated PowerEdge-based server building-blocks configured for use with PowerFlex. They are available with thousands of configuration options for customers who prefer to build their own environments.

PowerFlex software storage services can be deployed on recommended instances (with attached storage) in Amazon Web Services or Microsoft Azure. Only the EC and MG data layout as well as the Independent (disaggregated) architecture are supported. Where available, native asynchronous replication may be used to migrate data between cloud and on-premises PowerFlex systems, or to establish cloud-based BC/DR data protection schemes. This architecture is ideal for high-performance databases and application workloads.

An important component outside of PowerFlex that enables a flexible consumption model for Kubernetes is the PowerFlex CSI driver, developed as a part of the Dell Kubernetes strategy. After loading the CSI driver for PowerFlex into Kubernetes, it can be used to provision persistent volumes from the underlying PowerFlex storage resource. If the Kubernetes deployment is running low on PowerFlex storage resources, you can add PowerFlex storage nodes to increase the system capacity and performance.

The CSI driver connects the PowerFlex system and Kubernetes deployments. It is a storage broker which dynamically provisions volumes from PowerFlex through the PowerFlex API gateway to the Kubernetes cluster. Once the volume is available on PowerFlex, it is immediately mapped to the requesting pod. If a pod is destroyed or rescheduled, the CSI plug-in ensures that the volumes are remapped upon rescheduling of that pod. 

Customers running Kubernetes clusters on PowerFlex use the Dell Container Storage Modules (CSM), which extend the CSI driver capabilities. These modules:

• Provide enterprise storage capabilities to Kubernetes for cloud-native stateful applications.
• Reduce management complexity, so that developers can independently use enterprise storage with ease and automate daily operations.
• Extend storage functionality and capabilities beyond using the CSI driver alone.

These modules include replication, observability, authorization, application mobility, and resiliency.

PowerFlex supports multiple operating systems. PowerFlex is validated with the leading Kubernetes distributions as shown in the following figure:

Dell brings Kubernetes native disaster recovery for OpenShift by combining PowerFlex native asynchronous replication with the CSM Replication module. PowerFlex provides storage level, asynchronous data replication between independent clusters, where each cluster can act as both a source and a target. This bidirectional design enables flexible, geographically distributed architectures while ensuring resilient data protection for stateful workloads. Replication is handled at the array level, minimizing application impact and maintaining storage consistency.

The CSM Replication module extends the PowerFlex CSI driver to make this capability Kubernetes aware and declarative. Using Kubernetes CRDs, it orchestrates volume and volume group replication across clusters, supporting both planned migrations and unplanned disaster recovery without application changes. CSM replication coordinates failover and recovery workflows directly within the Kubernetes control plane and is deployed and managed through the CSM Operator. Together, PowerFlex and CSM Replication provide a simple, transparent, and enterprise-grade DR solution for OpenShift environments.

The CSM Resiliency module strengthens application availability and fault tolerance for Kubernetes workloads consuming PowerFlex persistent storage. It proactively improves storage behavior during node, pod, and infrastructure failures by enabling automated recovery actions and faster remediation. The module leverages Kubernetes health signals and PowerFlex storage capabilities to improve volume attachment reliability and reduce recovery time objectives. By coordinating failure handling across both the storage and orchestration layers, it minimizes application disruption and enhances platform stability. Dell delivers and manages the Resiliency module through the CSM Operator to ensure consistent behavior across OpenShift clusters.

Red Hat OpenShift Container Platform provides developers and IT organizations with a hybrid cloud application platform for deploying both new and existing applications on secure, scalable resources with minimal configuration and management overhead. The OpenShift Container Platform provides an enterprise-grade Kubernetes environment for building, deploying, and managing container-based applications, alongside virtualized workloads, across any data center where Red Hat Enterprise Linux is supported.

For more information, see Red Hat OpenShift Container Platform documentation.

OpenShift Assisted Installer provides a SaaS-based managed service to deploy OpenShift clusters in various configurations, including bare metal servers and on-premises nodes, from the Red Hat Hybrid Cloud Console. You do not require a dedicated bootstrap node before installing a cluster and you do not need to preinstall Red Hat CoreOS on the nodes. Cluster deployment parameters are populated within the web UI and nodes are booted with an ISO image that is downloaded to complete the cluster deployment.

For more information, see OpenShift Assisted Installer blog.

OpenShift Virtualization is an add-on to OpenShift Container Platform that allows you to run and manage VM workloads alongside container workloads. OpenShift Virtualization is based on upstream KubeVirt, which implements the deployment and management of VMs using Kubernetes constructs such as PVCs and pods. The following figure shows the KubeVirt components that are implemented in OpenShift Virtualization:

The KubeVirt components of OpenShift Virtualization are:

virt-controller

An operator that detects when new VM resources are created and creates the corresponding pod in which the VM runs. After the pod is scheduled to a node, the virt-controller provides the VM object with the node name and hands off to the virt-handler.

virt-handler

A daemon set running on every compute node, which performs all the steps that are required to move a VM object to the p state. This task is performed by referencing the VM specification and then creating a libvirtd domain on the VM pod. In the context of VMs, "p-state" refers to a processor's performance state, which dictates its operating frequency and voltage, allowing for dynamic adjustments to power consumption and performance based on workload demands. Virt-handler turns off the corresponding domain after the VM object is deleted.

virt-launcher

As the primary container running on every VM pod, this component creates the necessary groups and namespaces to host a VM after it is provided with a VM object to deploy from virt-handler. An instance of libvirtd hosted within virt-launcher is used to create the VM. After the VM is created, virt-launcher waits until the VM process is completed and then terminates itself. Virt-launcher also attempts to prevent preemptive shutdowns from Kubernetes to ensure that the VM exits before virt-launcher terminates.

libvirtd

An instance of libvirtd exists in every VM pod and is used by virt-launcher to manage the life cycle of the VM object.

For more information, see OpenShift virtualization architecture.

This solution describes the design and implementation of high availability and disaster recovery of VMs running on the OpenShift Virtualization platform with PowerFlex storage using PowerFlex CSM. The VMs are protected at the OCP cluster level using the PowerFlex CSM Resiliency module. Site-level recovery of the VMs is achieved using the PowerFlex CSM replication module.

This architecture is designed by creating two identical systems representing the source system and the target system. The source system is an active system with the VM workload, and the target system represents the DR system.

The design and implementation of the systems represented in the preceding diagram consist of the following components.

• Two OCP clusters, with one deployed on a source system and one on a target system. Each of these clusters is enabled with the OpenShift Virtualization feature. Both OCP clusters are deployed using the assisted installer with a configuration of three control nodes and three worker nodes. For the installation and recommended settings, see the Dell document Red Hat OpenShift Virtualization with Dell PowerFlex.
• Two PowerFlex storage-only clusters, with one cluster running on the source system and one on the target system. The PowerFlex clusters are built using the PowerFlex Manager GUI. Two PowerFlex management platforms are used for this purpose. A single PowerFlex protection domain and a single storage pool are configured on each PowerFlex cluster. The PowerFlex storage pool topology consists of four SDSs with 10 SSD disks in each of the four SDS nodes.
• Dell CSM is installed with the Replication and Resiliency modules. The PowerFlex CSI driver is installed with Replication and Resiliency sidecars enabled, for this purpose. The PowerFlex CSI driver enables dynamic provisioning and management of PowerFlex block storage for the stateful applications running on the OCP cluster. The CSM replication feature replicates the PVCs between source and target OCP clusters. The CSM resiliency feature provides cluster-level high availability for the pods during node failure.
• VMs are created on the source and the target OCP clusters using the OpenShift web console to validate this solution. The PostgreSQL instance is installed on these VMs, and the replication-enabled PVCs are used to create data and log mounts of the PostgreSQL databases.

The following diagram shows the network design used in this solution:

Each of the PowerFlex SO nodes and OCP nodes is attached with 2 x 25 GbE dual-port NICs. The node ports are trunked with the top-of-rack (TOR) switches. At the physical layer, two supported TOR switches are used for redundancy and load-balancing purposes. A peer link is configured on both TOR switches. VLANs are created to separate different traffic types on each node.

The VLANs configured for this solution are described in the following tables:

High availability for virtual machines running on the OpenShift Virtualization platform is enabled using the Dell Container Storage Modules (CSM) Resiliency feature, which provides node-level resiliency for VMs. Dell CSM replication capabilities provide system-level or site-level protection to support disaster recovery and minimize downtime during system-level or site-level failures. The detailed configuration and implementation steps for these capabilities are described in the following sections.

Dell PowerFlex CSI driver creates the persistent volumes (PVs) from Persistent Volume Claims (PVC) provided by users in their Kubernetes manifest YAML files. CSM Replication protects these PVs by replicating them at target volumes and managing operations such as replication synchronization and failover. It is important to note that CSM Replication maintains the same names for PVCs and PVs across the replication source and target. In addition, CSM creates the PowerFlex volumes at source system with the same name as the PVC name and at target system prefixes with the term replicated, making it even easier to manage and identify. However, keep in mind that the PowerFlex source and target volume IDs are unique to their system and do not match between source and target storage systems.

Although CSM Replication protects the PVs as explained above and provides an easy way to identify them on both source and target systems, this protection does not extend to the user application or VM deployment manifests. These YAML manifests and configuration files are stored outside the PVs and therefore are not replicated by PowerFlex CSM. As a result, if a disaster strikes and a replication failover operation takes place, the PVs will be available, but the manifests with the deployment instructions may not. It is therefore highly recommended that the user establish a process to store such files on an external NFS mount or a separately replicated volume to protect these files and make them available at the remote site if needed.

The following prerequisites must be met to implement PowerFlex CSM Resiliency and Replication on OpenShift Virtualization.

• Two PowerFlex storage-only clusters must be deployed with the PowerFlex replication enabled on them. These clusters must be in a healthy state and must be connected to each other by peering them in PowerFlex Manager.
• Two OpenShift clusters must be deployed to represent the source system and the target system for replication. Each of these clusters must be connected to the PowerFlex storage system.
• DNS and NTP are configured for name resolution and time synchronization across all OCP nodes and PowerFlex nodes.
• OpenShift Virtualization is enabled on both OCP clusters. Select Operators > Virtualization in Assisted Installer during the OCP cluster deployment.

  

  Note: OpenShift Virtualization can also be enabled using two additional methods—from the web console or from the CLI. See the Dell document Red Hat OpenShift Virtualization with Dell PowerFlex for details.
  
  
• The PowerFlex CSI driver is installed on both OCP clusters. Dell CSM Replication and Resiliency modules must be enabled during the installation. These steps are described in Appendix B - Dell CSM replication and resiliency. Verify the PowerFlex CSM pod status to ensure the CSM modules are running in both OCP clusters.

  

The Dell CSM Resiliency module is designed to make the application or VMs running on the OCP cluster resilient during node failures. This module is deployed as a side car to the PowerFlex CSI driver.

The Resiliency module detects the pod failure during the node failure, control plane failure, or network failure, cleans up the pod artifacts from the failed nodes, revokes the PV access from the failed node, and recreates the pod on a healthy node by reattaching the PV.

After the Resiliency module is enabled, a pod must be labeled to make it resilient. Identify the pod on which the OpenShift Virtualization VMs are running, then label the pod to make it resilient as shown in the following screenshot.

In OpenShift Virtualization, virtual machines are provisioned through a Kubernetes‑native workflow using either the OpenShift web console or declarative YAML manifests. During VM creation, compute resources, networking, and boot sources are defined, while storage is attached by referencing Persistent Volume Claims.

When using Dell PowerFlex CSM replicated PVCs, a storage class with replication policy is selected, triggering PowerFlex to provision synchronized volumes across multiple nodes. The PVC is then attached to the VM as a virtual disk, providing persistent storage with replication-based resiliency. This integration enables seamless VM lifecycle operations such as restart or migration without data loss, leveraging PowerFlex replication for fault tolerance.

To validate this solution, two PVCs are created using the replication-enabled PowerFlex storage class. PowerFlex CSI manages this storage class and provisions PowerFlex volumes at the back end.

The following figure shows the PVC creation in the OpenShift console using the replication-enabled storage class vxflexos-replication, Access mode RWO and Volume mode Block.

The PowerFlex CSI creates two PowerFlex volumes in the backend with the same name as the PV names. Verify this in PowerFlex Manager on the source system under the tab Block > Volumes.

The Dell replication controller manager detects this new PVC creation request in the replicated storage class and creates a DellCSIReplicationGroup (RG)—a Kubernetes Custom Resource to enforce the defined policy within the storage class. RG is a logical PowerFlex construct that defines how volumes are replicated, placed, and protected across failure domains.

The CSM Replication controller replicates these two PVCs to the target OCP cluster. The PVC and PV are created with the same name on the target OCP cluster.

In the backend, the PowerFlex volumes are replicated to the target PowerFlex cluster. Verify this in PowerFlex Manager on the target system.

In OpenShift Manager, create a VM by choosing Create > From InstanceType > rhel9 > U series (2 CPU, 4 GiB Memory).

Two disks are added using the PVCs created in the preceding section.

To create a VM disk, navigate to VM > Configuration > Storage > Disks > Add > Volume, then choose the replication-enabled PVC, Type as Disk, and Interface as VirtIO.

The deployment is validated in a Dell lab by performing a set of use cases on a VM running on the OpenShift Virtualization platform with PowerFlex software-defined infrastructure. A PostgreSQL instance is deployed on this VM. A database workload of 100 GB is created using pgbench for this purpose. Three use cases are performed to demonstrate high availability and disaster recovery of a VM using PowerFlex CSI Resiliency and Replication capabilities. The details are described in the following sections.

This use case validates VM level resiliency during a node failure in an OCP cluster. The node on which the workload VM pod is running is brought down manually. The virt-launcher pod is recreated and the workload VM is successfully rebooted on the next available OCP worker node.

This use case validates node-level high availability for virtual machines using the Dell CSM Resiliency module on OpenShift Virtualization. When the OpenShift node hosting vm1 entered maintenance mode, the platform automatically recreated the virt-launcher pod on a healthy node and reattached the PowerFlex-backed PVCs without manual intervention. Restarting the PostgreSQL instance after recovery confirms that the data remains intact, demonstrating dependable VM recovery and persistent data protection during node outages.

This use case demonstrates a failover from the source system to the target system and then runs reprotect to synchronize the data between the new source and the new target systems. The ocpcluster1 (source) is deployed with a PostgreSQL instance. A test1 database is created with a data size of 102 GB using pgbench.

This use case validates site-level disaster recovery for virtual machines using the Dell CSM Replication module with PowerFlex storage. After initiating a controlled failover from the source OpenShift cluster to the target cluster, the replicated PVCs were successfully promoted and attached to the target VM. Bringing up the PostgreSQL instance on the target system confirms that the database and all associated data are fully preserved. After the successful failover, a reprotect RG is triggered to enable replication from the new source to the new target. This demonstrates that Dell CSM replication enables dependable VM failover with consistent data availability during site-level outages.

This use case performs a failback to the original source system to set the replication in the original direction. Before setting the replication direction to the original path, confirm that the replication group (RG) status is healthy and the replication between the source and the target systems is functioning correctly.

This use case validates the failback operation to the original source system using the Dell CSM replication module with PowerFlex storage. After completing reverse replication and initiating failover back to the original OpenShift cluster, the replicated PVCs were successfully promoted and reattached to the original VM. Restarting the PostgreSQL instance confirmed that all data, including updates made during the DR operation, was preserved. This demonstrates that Dell CSM replication supports dependable VM failback while maintaining data consistency and continuity across sites. The Reprotect RG operation is performed to synchronize the data and resume replication between the original source system and the original target system.

This solution delivers a unified, storage-native HA and DR framework for virtual machines running on OpenShift Virtualization with Dell PowerFlex. By combining policy-driven resiliency and replication through Dell CSM, it enables high availability, predictable disaster recovery, and operational simplicity across sites. The validated architecture and workflows provide a technical reference for designing and implementing VM data protection in hybrid OpenShift environments. To discover more, contact your Dell representative.

This section provides the commands used in this solution during the resiliency and replication operations.