FlashStack with Nutanix: A Validated Exit Path from VMware

FlashStack with Nutanix is a converged infrastructure architecture jointly validated and documented by Cisco, Pure Storage, and Nutanix. It represents a fundamental shift from the traditional FlashStack approach — instead of VMware vSphere as the hypervisor layer, it uses Nutanix AHV while maintaining the proven Cisco UCS compute and Pure Storage foundation.

The architecture combines Cisco UCS compute (X-Series, C-Series, or B-Series server families) managed via Cisco Intersight in Intersight Managed Mode (IMM), Pure Storage FlashArray as external all-flash storage connected via NVMe/TCP or iSCSI, and Nutanix Cloud Platform software running AOS (Acropolis Operating System) with Nutanix AHV as the hypervisor.

What makes this architecture significant is the joint validation process. The three vendors tested this configuration together and published a Cisco Validated Design (CVD) covering design, deployment, and operational guidance. This means TAC at all three vendors understands the stack, interoperability has been verified across firmware and software versions, and there's a predictable upgrade path with coordinated version combinations.

The classic FlashStack architecture — Cisco UCS + Pure Storage + Cisco Nexus + VMware vSphere — has been a dominant enterprise converged infrastructure stack for over a decade. Tens of thousands of organizations run it successfully, with proven operational patterns and deep institutional knowledge.

When Broadcom restructured VMware licensing in 2024, these organizations faced three difficult options. Option A was renewing vSphere at dramatically higher cost, buying capabilities bundled in VCF that many organizations don't need. Option B meant replacing the entire stack — new hardware, new storage, new hypervisor — which represents massive capital expenditure and operational risk. Option C was replacing only the hypervisor while keeping the hardware that's working.

FlashStack with Nutanix is Option C, now jointly validated by all three vendors. Organizations running FlashStack today can keep their existing Cisco UCS hardware (X-Series, C-Series, B-Series back to generation M5), keep their existing Pure FlashArray (if on a supported model and Purity OS version), add Nutanix Cloud Platform software and AHV hypervisor, and retire VMware vSphere licensing.

The result is the same hardware foundation with lower software costs and a vendor-validated architecture that maintains support relationships across all components.

The FlashStack with Nutanix architecture consists of four primary layers, each with specific responsibilities and integration points.

The compute foundation uses Cisco UCS in one of three form factors. UCS X-Series modular is recommended for new builds — it's modern, chassis-based, and Intersight-native. UCS C-Series rack provides rack-mount options, either standalone or Intersight-managed. UCS B-Series blade systems are still supported for existing environments.

All systems are managed via Cisco Intersight in Intersight Managed Mode (IMM), providing cloud-operated lifecycle management. Service Profile templates (called Server Profiles in IMM) define vNIC/vHBA configuration, firmware policy, and boot policy per group of servers, ensuring consistent configuration across the compute layer.

Pure FlashArray provides all-flash external storage, connected via NVMe/TCP or iSCSI over standard Ethernet. NVMe/TCP delivers NVMe protocol performance over standard IP network infrastructure — no Fibre Channel HBAs or FC switches required.

Pure's Evergreen model supports independent controller upgrades without impacting active workloads, enabling non-disruptive maintenance. Volume groups and host mappings are defined per UCS host (Nutanix compute node), with inline deduplication and compression providing data reduction at the FlashArray level.

Nutanix AHV runs on UCS compute nodes and connects to Pure storage via NVMe/TCP or iSCSI. AHV is a KVM-based hypervisor included in the Nutanix AOS license at no additional hypervisor cost — a key differentiator from VMware's per-socket licensing model.

Prism Central provides VM management, policy enforcement, migration scheduling, and protection domain management across the AHV cluster.

While original FlashStack architectures typically used Cisco Nexus switching, FlashStack with Nutanix in IVI's AIM context leverages Arista Ethernet fabric. This includes Arista leaf-spine architecture with dedicated storage VLANs, jumbo frame support, and deep-buffer leaf switches optimized for storage traffic patterns.

FlashStack with Nutanix requires coordination across three separate management planes, each with distinct responsibilities but integrated through open APIs and unified monitoring.

Cisco Intersight owns UCS hardware lifecycle, server health monitoring, firmware compliance tracking, service profile management, and Intersight Actions for automated remediation. This provides centralized UCS management across all compute nodes in the cluster.

Nutanix Prism Central manages VM lifecycle operations including create, modify, and delete functions, AHV host management, cluster health monitoring, storage policy for Nutanix-native storage elements, and protection domains with replication capabilities.

Pure1, Pure Storage's cloud management platform, handles FlashArray health monitoring, capacity management, performance analytics, replication status tracking, and Evergreen hardware management for non-disruptive upgrades.

IVI serves as the integration specialist for this three-vendor architecture. While each vendor's TAC supports their own component, the integration layer — how the three components work together operationally — is IVI's domain of expertise.

During design and deployment, IVI provides architecture assessment to right-size compute and storage for target workloads, FlashStack configuration including UCS service profiles and FI uplink design, Pure volume groups and AHV host definitions, NVMe/TCP or iSCSI host connectivity setup, network integration with Arista fabric design for storage VLANs, Nutanix cluster formation and Prism Central registration, Pure FlashArray host definition and target configuration, and integration validation through storage path performance testing before workload deployment.

For VMware exit scenarios, IVI manages Nutanix Move for live VM migration from vSphere to AHV, per-workload migration wave planning, post-migration validation and stabilization, and vSphere decommission planning and execution.

Ongoing operations through Aegis Managed Services include unified monitoring via Aegis PM, firmware and OS lifecycle management across all three platforms, 24x7 incident response with single point of contact, and vendor TAC management across Cisco, Pure, and Nutanix relationships.

FlashStack with Nutanix serves two primary market segments with different motivations and requirements.

The primary target consists of existing FlashStack customers — organizations already running Cisco UCS + Pure Storage + VMware vSphere who need a cost-effective path to exit VMware hypervisor licensing without replacing proven, working hardware. This represents the most compelling use case because it protects existing investment, swaps only the hypervisor layer, and provides a jointly validated architecture with maintained vendor support.

The secondary target includes greenfield AIM buyers who want a validated reference design. These organizations are building new environments and prefer a three-vendor validated CVD as their starting point rather than designing a custom architecture from scratch.

This architecture is not a fit for organizations running generic x86 servers from Dell or HP without UCS — while they can still run Nutanix AHV with Pure storage, it wouldn't constitute a FlashStack architecture. It's also not suitable for organizations that want full HCI with no external storage, who should consider Nutanix HCI or UCS-based Nutanix NX equivalents instead. Finally, organizations with NSX-T dependency where network virtualization migration is the primary complexity should note that FlashStack with Nutanix addresses compute, storage, and hypervisor layers but not SDN migration requirements.

Five key questions determine whether FlashStack with Nutanix is the right architectural choice for your organization.

First, are you currently running FlashStack with Cisco UCS, Pure Storage, and VMware? If yes, FlashStack with Nutanix represents the direct evolution path with lowest migration risk and capital cost, though hardware compatibility will determine project scope.

Second, what generation is your UCS hardware? X-Series, B-Series M5 and newer, and C-Series M5 and newer are likely supported, while older generations require verification against Nutanix and Cisco compatibility matrices. If hardware is not supported, evaluate whether hardware refresh plus FlashStack with Nutanix is still the right path versus a completely new architecture.

Third, what generation and model is your Pure FlashArray? Current FlashArray models (//X, //C, //XL) are fully supported with NVMe/TCP connectivity. Older models may require iSCSI connectivity rather than NVMe/TCP, and some models may be approaching end-of-life — verify against Pure compatibility matrices.

Fourth, what is your Broadcom renewal timeline? This drives project timeline constraints. FlashStack with Nutanix migration can typically be executed in 12-20 weeks for a standard environment, but complex configurations may require longer planning cycles.

Fifth, do you have VMware NSX-T deployed? If yes, hypervisor migration and network virtualization migration become separate workstreams. Plan accordingly — NSX-T replacement is not included in FlashStack with Nutanix scope and may require additional architecture decisions.

FlashStack with Nutanix migration involves several technical and operational considerations that determine project complexity and timeline.

Live migration from VMware vSphere to Nutanix AHV is handled through Nutanix Move, which provides cross-hypervisor VM migration capabilities. This allows for phased migration approaches where workloads can be moved in planned waves, reducing operational risk and allowing for validation at each stage.

Network configuration requires careful planning, particularly around storage VLANs and NVMe/TCP or iSCSI connectivity. The migration from Cisco Nexus to Arista fabric (in IVI AIM context) may require network architecture updates, though this can often be coordinated with the hypervisor migration to minimize disruption windows.

Management workflow changes are significant — operations teams need training on Nutanix Prism Central for VM management, understanding the integration points between Intersight, Prism Central, and Pure1, and adapting incident response procedures for the three-vendor architecture.

Licensing considerations include VMware license retirement timing, Nutanix AOS licensing requirements, and ensuring Cisco Intersight IMM licensing is in place for UCS management. The financial model shifts from VMware's per-socket licensing to Nutanix's per-node licensing, typically resulting in lower total software costs.