Skip to main content

Step-by-Step Guide: Deploying a 3-Node Nutanix Cluster (AHV/AOS) on Cisco UCS FIs & Deploying Prism Central on ESXi

  • August 3, 2026
  • 1 reply
  • 6 views

Moaz Ahmed Abdelghany
Forum|alt.badge.img+1

Deploying a Nutanix cluster running Nutanix AHV and AOS on Cisco UCS hardware with Fabric Interconnects (FIs) provides a resilient, high-performance hyperconverged infrastructure. Combining Nutanix with UCS Fabric Interconnects requires proper VLAN configurations, network mapping, and a seamless Foundation process.

Furthermore, managing multi-cluster environments or advanced management services often requires deploying Prism Central. In this guide, we will walk through the end-to-end architecture: deploying a 3-node Nutanix AHV/AOS cluster, followed by deploying Prism Central as a VM on an ESXi host, registering Prism Element to it, and configuring network settings via the GUI.

Architecture Overview & Prerequisites

Architecture Topology

  • Cluster Nodes: 3 x Cisco UCS C-Series Rack Servers (or UCS B-Series Blades)

  • Control Plane: 2 x Cisco UCS Fabric Interconnects (FI 6400 / 6300 Series) in High Availability mode (FI-A and FI-B)

  • Storage & Hypervisor: Nutanix AOS & Nutanix AHV

  • Management Engine: Prism Central deployed as a standalone Virtual Machine on an external/existing VMware ESXi host.

  • Fabric Connections: Dual-homed connectivity (each node has redundant uplinks split between FI-A and FI-B).

   +-------------------------------------------------------+
| Core Switches / Upstream |
+---------------------------+---------------------------+
|
+-----------+-----------+
| |
+-------+-------+ +-------+-------+
| Cisco UCS FI-A| | Cisco UCS FI-B|
+-------+-------+ +-------+-------+
| |
+-----------+-----------+-----------+-----------+
| | |
+------+--------+ +------+--------+ +------+--------+
| Nutanix Node 1| | Nutanix Node 2| | Nutanix Node 3|
+---------------+ +---------------+ +---------------+

Pre-requisites Checklist

  1. UCS Manager / Intersight: Fully configured Fabric Interconnect pair with active uplinks to upstream switches.

  2. IP Address Allocation:

    • Out-of-band IPMI / Cisco IMC IPs for each node.

    • Host IPs (AHV Hypervisor) for 3 nodes.

    • Controller VM (CVM) IPs for 3 nodes.

    • Cluster Virtual IP (VIP) for Nutanix Prism Element.

    • Prism Central VM IP (for deployment on ESXi).

    • Nutanix Foundation VM / Deployment Laptop IP.

  3. VLANs Configured on UCS FI:

    • Management VLAN (for AHV, CVM, and IPMI/CIMC).

    • Data / Storage VLANs.

    • VM Guest Networks.

Step 1: Cisco UCS Fabric Interconnect Configuration

Before starting the Foundation process, the UCS network profiles must be correctly provisioned:

  1. Create MAC Address Pools: Define redundant MAC pools for vNICs across FI-A and FI-B.

  2. Configure VLANs:

    • In UCSM / Intersight, navigate to LAN -> VLANs and create your required Nutanix Management and Production VLANs.

    • Mark the Nutanix Management VLAN as Native if untagged traffic is used during initial deployment.

  3. Configure Service Profiles / VNIC Templates:

    • Assign two physical vNICs per server: vNIC_A (bound to FI-A) and vNIC_B (bound to FI-B).

    • Set MTU to 9000 (Jumbo Frames) on UCS QOS policies for optimal AOS storage performance.

Step 2: Preparing Nutanix Foundation & Environment Setup

To ensure uninterrupted communication (IPv6 discovery, SSH, TFTP, and imaging protocols) during the Foundation process, proper firewall management is essential:

  1. Disable Host Laptop Firewall: Completely turn off the local firewall (Windows Firewall or macOS Firewall) on your management laptop to allow unhindered traffic between your machine, the Foundation VM, and the target nodes.

  2. Disable Foundation VM Firewall: Ensure the internal OS firewall inside the Nutanix Foundation VM is stopped:

    Bash
     
    sudo systemctl stop firewalld
    sudo systemctl disable firewalld
  3. Connect your deployment laptop/VM to the same flat L2 broadcast domain as the CIMC/IPMI ports of the 3 Cisco nodes.

Step 3: Cluster Discovery, Imaging & Prism Element Initialization

  1. Launch Discovery: Open Nutanix Foundation and run a scan to discover the 3 Cisco UCS nodes via IPv6 / MAC address.

  2. Select Software Stack: Select Nutanix AOS and Nutanix AHV as the target deployment stack.

  3. Network & Node Configuration:

    • Assign static IPs for CIMC, AHV Host, and CVM for all 3 nodes.

    • Provide Subnet Mask, Gateway, and Management VLAN ID.

  4. Create Cluster: Check "Create Cluster", set the Cluster Name, Redundancy Factor (RF2), Cluster Virtual IP (VIP), and DNS/NTP servers.

  5. Start Foundation Process: Click Start Foundation. Once finished, the 3-node AHV/AOS cluster will be running.

  6. Access Prism Element: Open https://<Cluster-Virtual-IP>:9440, set a new admin password, and accept the EULA.

Step 4: Deploying Prism Central on ESXi & Registering Prism Element

Once Prism Element is up and running, proceed with deploying Prism Central as a VM on an external VMware ESXi host:

1. Prism Central VM Deployment (OVA on ESXi)

  1. Download the Prism Central OVA package for ESXi from the Nutanix Portal.

  2. Log in to your VMware ESXi / vCenter server.

  3. Deploy the OVF/OVA template, allocating the required CPU, RAM, and Storage resources (Small/Medium size based on environment requirements).

  4. Power on the Prism Central VM and complete initial IP/Network setup via console or Web UI wizard.

2. GUI Network & Cluster Configuration in Prism Central

  1. Access Prism Central GUI at https://<Prism-Central-IP>:9440.

  2. Log in with default credentials (admin), set a new password, and navigate to Prism Central Settings.

  3. Network Configuration (GUI):

    • Configure Virtual Switches / Subnets under Network Settings.

    • Assign static management IP, subnet mask, gateway, and DNS/NTP servers directly through the GUI interface.

3. Registering Prism Element to Prism Central

  1. Open Prism Element (https://<Cluster-Virtual-IP>:9440).

  2. Navigate to Gear Icon (Settings) -> Prism Central Registration.

  3. Click Register, enter the Prism Central IP Address and admin credentials.

  4. Confirm registration. Once complete, your 3-node Nutanix AHV cluster is fully integrated and managed via Prism Central!

1 reply

Allan Solomon Mejia
Forum|alt.badge.img+1

Excellent walkthrough ​@Moaz Ahmed Abdelghany . Deploying Nutanix on Cisco UCS with Fabric Interconnects demonstrates that successful implementations depend just as much on careful planning as they do on the deployment steps themselves.

From my experience, there are a few best practices that make a significant difference in production environments:

  • Validate firmware, UCS, AHV, AOS, and hardware compatibility before deployment to avoid unexpected interoperability issues.
  • Ensure DNS, NTP, VLANs, MTU, and IP addressing are fully validated before bringing the cluster online. Small configuration issues here can consume hours of troubleshooting later.
  • If Prism Central is initially deployed on an external ESXi environment for cluster provisioning, plan to migrate to a Prism Central instance running on the Nutanix cluster for long-term operations, which is also Cisco's recommended approach.
  • Perform a post-deployment health check by running NCC and validating cluster resiliency, storage health, and failover scenarios before handing the environment over to operations.

One thing I've found is that the deployment itself is usually the easiest part. Building a platform that's easy to operate, patch, scale, and recover over the next five years is where good architecture really pays off. Great job documenting the process. This will be a valuable reference for anyone deploying Nutanix on Cisco UCS.