Deploying Nexus Dashboard 4.2 via VMware vCenter: Step-by-Step

Overview

This guide covers deploying a Nexus Dashboard 4.2 cluster using VMware vCenter. If you’re deploying directly on ESXi without vCenter, see the companion post for the direct ESXi deployment procedure.

Before starting, ensure you’ve completed all prerequisites — see the prerequisites post for full details on node sizing, storage requirements, and critical restrictions (DRS, vMotion, snapshots).

Step 1 — Download the OVA Image

Navigate to the Cisco Software Download page:

https://software.cisco.com/download/home/286327743/type/286328258/

Select your Nexus Dashboard release version and download the OVA image:

nd-dk9.<version>.ova

You can either download the OVA locally and select it as a local file during deployment, or host it on an internal web server and provide the URL during the OVA deployment wizard.

Step 2 — Launch the Deploy OVF Template Wizard

1. Log in to VMware vCenter
2. Right-click the ESXi host where you want to deploy the first node
3. Select Deploy OVF Template…

The following steps reference vSphere Client 7.0. Screen layout may differ slightly in other versions.

Step 3 — Select OVF Template Source

In the Select an OVF template screen:

• If hosting on a web server: select URL and provide the full URL to the OVA
• If local file: select Local file and click Choose Files, then select the downloaded OVA

Click Next.

Step 4 — Name and Location

In the Select a name and folder screen:

1. Enter a VM name — use a consistent naming convention, for example:

   nd-ova-node1
   nd-ova-node2
   nd-ova-node3
   

2. Select the datacenter or folder location for the VM
3. Click Next

Step 5 — Select Compute Resource

In the Select a compute resource screen:

1. Select the vCenter datacenter and the ESXi host for this node
2. Click Next

Deploy each node on a different ESXi host for hardware redundancy.

Step 6 — Select the Node Profile

In the Configuration screen, select the node profile:

• App — 16 vCPUs, 64 GB RAM, 500 GB storage. Covers most deployments.
• Data — 32 vCPUs, 128 GB RAM, 3 TB SSD. Required for high-scale or specific telemetry features.

Select the appropriate profile based on your capacity planning and click Next.

Step 7 — Storage Configuration

In the Select storage screen — this step has several critical settings:

1. Select the datastore for the VM (use a unique datastore per node where possible)
2. ✅ Check “Disable Storage DRS for this virtual machine” — this is mandatory
3. From the Select virtual disk format dropdown, choose Thick Provisioning Lazy Zeroed
4. Click Next

Do not use Thin Provisioning. Nexus Dashboard requires thick-provisioned disks for consistent I/O performance.

Step 8 — Network Mapping

In the Select networks screen, map the two required Nexus Dashboard networks:

Both interface pairs use bonding for high availability. Select the appropriate port groups for each and click Next.

Management and data networks must be in different subnets.

Step 9 — Customize Template

In the Customize template screen:

1. Enter and confirm the Password — this becomes the rescue-user SSH password for each node
2. Enter the Management Network IP address and netmask for this specific node
3. Enter the Management Network gateway
4. Click Next

⚠️ You must use the same password for all nodes in the cluster. Different passwords will cause cluster creation to fail.

Step 10 — Review and Deploy

In the Ready to complete screen, verify all settings are correct and click Finish to begin deploying the first node.

Step 11 — Deploy the Remaining Nodes

Repeat Steps 2–10 for the second and third primary nodes.

• You do not need to wait for the first node to finish deploying before starting the next — deploy all three simultaneously
• Each node needs a unique management IP address
• All nodes must use the same rescue-user password

For single-node clusters (lab/non-production), skip this step.

Step 12 — Disable VMware Tools Time Synchronization

Before powering on the VMs, disable periodic time sync on each node:

1. Right-click the node’s VM → Edit Settings
2. Select the VM Options tab
3. Expand VMware Tools
4. Uncheck “Synchronize time periodically”
5. Click OK

Repeat for each ND node. Then power on all VMs.

Step 13 — Access the Bootstrap GUI

Once the VMs are running, open a browser and navigate to:

https://<node-mgmt-ip>

You can use any one of the three nodes to complete the cluster bootstrap — you don’t need to log into all three.

Enter the rescue-user password you set during deployment and click Login.

Step 14 — Cluster Bringup Wizard

The Cluster Bringup wizard guides you through the remaining configuration. Key steps:

Basic Information

• Cluster Name — must follow RFC-1123 naming (lowercase alphanumeric, hyphens allowed)
• Implementation Type — choose LAN or SAN depending on your use case

Configuration Page

• Add DNS servers (required — ND will fail to deploy without valid DNS)
• Add NTP servers (required)
• Optional: configure NTP authentication using MD5, SHA, or AES128CMAC keys
• Optional: configure a proxy server for clusters without direct internet access

The proxy must allow these URLs if configured:

svc.intersight.com
svc-static1.intersight.com
svc-static1.ucs-connect.com

Node Details

For each node, configure:

1. Cluster Connectivity — choose L2 if all nodes are on the same subnet, or BGP for L3-adjacent nodes
2. Click Edit next to each node and provide:
   • Node Name (becomes the hostname — must follow RFC-1123)
   • Node Type: Primary for all three initial nodes
   • Data Network IP address, netmask, gateway
   • Optionally: VLAN ID (leave blank if not required)
   • If BGP is enabled: ASN and BGP peer details

If you need IPv6, you must configure it during bootstrap — it cannot be changed later without redeploying the cluster.

Persistent IPs

Persistent IP addresses are used for telemetry and service continuity. Minimum requirements:

• 1-node cluster: 3 persistent IP addresses on the data network
• 3-node cluster: 5 persistent IP addresses on the data network

Add these during the Persistent IPs page of the wizard.

Step 15 — Monitor Cluster Formation

Click Save and Continue to begin cluster formation. Progress is displayed in the UI for each node individually.

Cluster formation can take up to 60 minutes or more depending on the number of nodes. If progress appears stalled, manually refresh the browser — this is normal.

Step 16 — Verify Cluster Health

Once the cluster is up, SSH into any node and check:

$ acs health

Expected output while converging:

k8s install is in-progress
k8s services not in desired state - [...]
k8s: Etcd cluster is not ready

When healthy:

All components are healthy

If a node gets stuck at deploy base system services after a power cycle, run acs reboot clean on the affected node to resolve etcd issues.

You can also check the Anomaly Level on the Home > Overview page in the GUI — a healthy cluster will show no critical anomalies.

What’s Next

• Connect ND to Cisco Intersight for visibility and TAC log collection
• Create Fabrics and Fabric Groups to onboard your ACI or NX-OS fabric
• Install Nexus Dashboard services (NDFC, NDI, NDO) from the Admin > Services page