Deploy a fully functional VMware Cloud Foundation 9.1 environment on one, two or three physical ESX hosts, optimized for development and lab use. This setup enables users to explore the full capabilities of VCF 9.1, ranging from Fleet Management, Self-Service Automation with Multi-Tenancy to running modern container and AI workloads, all while using minimal compute and storage resources.
📒 This deployment does not use Nested ESX and instead runs directly on physical hosts, in contrast to the VCF Holodeck solution.
- Changelog
- Bill of Materials (BOM)
- Prerequisites
- Installation
- Post Installation
- Additional Blog Resources
- 05/28/2026
- Initial Release
- VMware Cloud Foundation (VCF) 9.1.0.0
- VCF Installer 9.1.0
- VCF 9.1.0 Binaries using VCF Offline Depot
- ~275GB Memory (highly recommend enabling NVMe Tiering with higher endurance drives for improved performance)
- Single Host Deployment ThinkStation P5
- Dual or Triple Host Deployment Minisforum MS-A2
📒 The above BOM is just an example. You can certainly swap out full systems and/or components that you might already have or prefer alternatives. Just know that you are responsible for adjusting any configuration that may differ from the referenced BOM.
-
Minimum 5 VLANs (e.g. 30, 40, 50, 60, 70) for VCF Fleet Deployment
- VLAN 30 - Management
- VLAN 40 - vMotion
- VLAN 50 - vSAN
- VLAN 60 - ESX/NSX Edge TEP
- VLAN 70
- Tier 0 Uplink for Centralized Transit Gateway (Optional)
- External Connectivity for Distributed Transit Gateway (Optional)
-
Here are the IP addresses and DNS entries that you will need for initial setup (NSX Edge/VNA and vSphere Supervisor are optional)
| Hostname | FQDN | IP Address | Function |
|---|---|---|---|
| esx01 | esx01.vcf.lab | 172.30.0.10 | Physical ESX-1 Server |
| esx02 | esx02.vcf.lab | 172.30.0.11 | Physical ESX-2 Server |
| esx03 | esx03.vcf.lab | 172.30.0.12 | Physical ESX-3 Server |
| sddcm01 | sddcm01.vcf.lab | 172.30.0.18 | VCF Installer / SDDC Manager |
| vc01 | vc01.vcf.lab | 172.30.0.19 | vCenter Server for Management Domain |
| vcf01 | vcf01.vcf.lab | 172.30.0.20 | VCF Operations |
| vcf-msr01 | vcf-msr01.vcf.lab | 172.30.0.21 | VCF Management Services Runtime |
| vcf-flt01 | vcf-flt01.vcf.lab | 172.30.0.22 | VCF Fleet Components FQDN |
| vcf-int01 | vcf-int01.vcf.lab | 172.30.0.23 | VCF Instance Components FQDN |
| vcf-lic01 | vcf-lic01.vcf.lab | 172.30.0.24 | VCF License Server |
| vcf-log01 | vcf-log01.vcf.lab | 172.30.0.26 | VCF Operations for Logs (Optional) |
| vcf-idb01 | vcf-idb01.vcf.lab | 172.30.0.27 | VCF Identity Broker |
| 172.30.0.32/28 | 172.30.0.33-172.30.0.46 | VCF Management Services Runtime Node IP Pool | |
| nsx01 | nsx01.vcf.lab | 172.30.0.48 | NSX Manager VIP for Management Domain |
| nsx01a | nsx01a.vcf.lab | 172.30.0.49 | NSX Manager for Management Domain |
| edge01a | edge01a.vcf.lab | 172.30.0.50 | NSX Edge 1a for Management Domain (Optional) |
| edge01b | edge01b.vcf.lab | 172.30.0.51 | NSX Edge 1b for Management Domain (Optional) |
| vna01a | vna01a.vcf.lab | 172.30.0.52 | NSX VNA 1a for Management Domain |
| vna01b | vna01b.vcf.lab | 172.30.0.53 | NSX VNA 1b for Management Domain (Optional) |
| 172.30.0.54 | NSX Edge VIP (Optional) | ||
| vcf-proxy01 | vcf-proxy01.vcf.lab | 172.30.0.55 | VCF Operations Proxy Collector |
| vcf-asr01 | vcf-asr01.vcf.lab | 172.30.0.56 | VCF Automation Services Runtime |
| auto01 | auto01.vcf.lab | 172.30.0.57 | VCF Automation |
| 172.30.0.64/29 | 172.30.0.65-172.30.0.70 | VCF Automation Services Runtime Node IP Pool | |
| sv01 | sv01.vcf.lab | 172.30.0.80-172.30.0.85 | vSphere Supervisor Node IP Pool |
📒 For Distributed Transit Gateway (DTGW) configuration, we can optimize the VLAN 70 network and carve it up into smaller /26 networks rather than consuming an entire block when enabling vSphere Supervisor
- 172.30.70.0/26
- 172.30.70.64/26
- 172.30.70.128/26
- 172.30.70.192/26
-
Update your hardware to the latest vendor firmware
- For MS-A2 owners, please follow these instructions for firmware
- Apply these these instructions for network optimizations if you are doing a manual installation (already incorporated into ESX kickstart examples)
-
Set up a VCF Offline Depot using the new
VCF Download Toolby following the Broadcom documentation
After downloading the required metadata/binaries, you should have a directory structure like the following:
PROD
├── COMP
│ ├── DEPOT_SERVICE
│ │ ├── configuration-schema-vcf-fleet-depot-9.1.0.0.25371105.yaml
│ │ ├── depot-manifest-vcf-fleet-depot-9.1.0.0.25371105.yaml
│ │ ├── vcf-fleet-depot-9.1.0.0.25371105.tgz
│ │ └── vcf-fleet-depot-plugin-9.1.0.0.25371105.tgz
│ ├── ESX_HOST
│ │ └── VMware-VMvisor-Installer-9.1.0.0.25370933.x86_64.iso
│ ├── NSX_T_MANAGER
│ │ ├── nsx-unified-appliance-9.1.0.0.25318227.ova
│ │ └── VMware-NSX-T-9.1.0.0.25318227.vlcp
│ ├── SDDC_MANAGER_VCF
│ │ ├── Compatibility
│ │ │ └── VmwareCompatibilityData.json
│ │ └── VCF-SDDC-Manager-Appliance-9.1.0.0.25371088.ova
│ ├── TELEMETRY_ACCEPTOR
│ │ ├── configuration-schema-telemetry-acceptor-9.1.0.0.25181946.yaml
│ │ ├── depot-manifest-telemetry-acceptor-9.1.0.0.25181946.yaml
│ │ ├── telemetry-acceptor-9.1.0.0.25181946.tgz
│ │ └── telemetry-acceptor-plugin-9.1.0.0.25181946.tgz
│ ├── VCENTER
│ │ └── VMware-VCSA-all-9.1.0.0.25370922.iso
│ ├── VCF_FLEET_LCM
│ │ ├── configuration-schema-vcf-fleet-lcm-9.1.0.0.25371109.yaml
│ │ ├── depot-manifest-vcf-fleet-lcm-9.1.0.0.25371109.yaml
│ │ ├── vcf-fleet-lcm-9.1.0.0.25371109.tgz
│ │ └── vcf-fleet-lcm-plugin-9.1.0.0.25371109.tgz
│ ├── VCF_LICENSE_SERVER
│ │ └── Vcf-License-Server-9.1.0.0.25346031.ova
│ ├── VCF_OPS_CLOUD_PROXY
│ │ └── Operations-Cloud-Proxy-9.1.0.0.25346033.ova
│ ├── VCF_SALT
│ │ ├── configuration-schema-salt-9.1.0.0.25346036.yaml
│ │ ├── depot-manifest-salt-9.1.0.0.25346036.yaml
│ │ ├── salt-9.1.0.0.25346036.tgz
│ │ └── salt-plugin-9.1.0.0.25346036.tgz
│ ├── VCF_SALT_RAAS
│ │ ├── configuration-schema-salt-raas-9.1.0.0.25346036.yaml
│ │ ├── depot-manifest-salt-raas-9.1.0.0.25346036.yaml
│ │ ├── salt-raas-9.1.0.0.25346036.tgz
│ │ └── salt-raas-plugin-9.1.0.0.25346036.tgz
│ ├── VCF_SDDC_LCM
│ │ ├── configuration-schema-vcf-sddc-lcm-9.1.0.0.25371107.yaml
│ │ ├── depot-manifest-vcf-sddc-lcm-9.1.0.0.25371107.yaml
│ │ ├── vcf-sddc-lcm-9.1.0.0.25371107.tgz
│ │ └── vcf-sddc-lcm-plugin-9.1.0.0.25371107.tgz
│ ├── VCF_SERVICE_VCD_MIGRATION_BACKEND
│ │ ├── configuration-schema-vcd-migrator-9.1.0.0.25370929.yaml
│ │ ├── depot-manifest-vcd-migrator-9.1.0.0.25370929.yaml
│ │ ├── vcd-migrator-9.1.0.0.25370929.tgz
│ │ └── vcd-migrator-plugin-9.1.0.0.25370929.tgz
│ ├── VIDB
│ │ ├── configuration-schema-vidb-9.1.0.0.25368698.yaml
│ │ ├── depot-manifest-vidb-9.1.0.0.25368698.yaml
│ │ ├── vidb-9.1.0.0.25368698.tgz
│ │ └── vidb-upgrade-plugin-9.1.0.0.25368698.tgz
│ ├── VRA
│ │ ├── configuration-schema-vcfa-bundle-9.1.0.0.25370929.yaml
│ │ ├── depot-manifest-vcfa-bundle-9.1.0.0.25370929.yaml
│ │ ├── vcfa-bundle-9.1.0.0.25370929.tar
│ │ └── vcfa-plugin-9.1.0.0.25370929.tgz
│ ├── VROPS
│ │ └── Operations-Appliance-9.1.0.0.25346025.ova
│ └── VSP
│ ├── configuration-schema-vmsp-platform-9.1.0.0.25370367.yaml
│ ├── depot-manifest-vmsp-platform-9.1.0.0.25370367.yaml
│ ├── vcf-services-platform-template-9.1.0.0.25370367.ova
│ ├── vmsp-cli-9.1.0.0.25370367.tar.gz
│ ├── vmsp-platform-9.1.0.0.25370367.tar
│ └── vmsp-plugin-9.1.0.0.25370367.tgz
├── metadata
│ ├── manifest
│ │ └── v1
│ │ └── vcfManifest.json
│ └── productVersionCatalog
│ └── v1
│ ├── productVersionCatalog.json
│ └── productVersionCatalog.sig
└── vsan
└── hcl
├── all.json
└── lastupdatedtime.json
27 directories, 56 files
You can host the VCF Offline Depot using a traditional HTTP Web Server (HTTPS is NOT required as the automation in 9.1 will support HTTP). Alternatively, you can simply use Python to serve up the directory (see this blog post) or even a Synology (see this blog post).
-
Create a bootable ESX installer with the ESX ISO (VMware-VMvisor-Installer-9.1.0.0.25370933.x86_64.iso) using UNetbootin.
-
We will perform a scripted installation of ESX (aka ESX Kickstart) to reduce the number of manual post-installation steps.
-
Edit the relevant Kickstart Files and replace the following values with your own desired configurations.
💡 To simplify the deployment of multiple ESX hosts using a single USB drive, you can create a custom UEFI boot menu for ESX, allowing you to select specific ESX kickstart configuration files.
📒 To identify the NVMe device label for the ESX installation (e.g. --disk=<ID>) and NVMe tiering device (e.g. NVME_TIERING_DEVICE=), boot the ESX installer initially, switch to the shell console (ALT+F1), and log in as root with a blank password (just press Enter). Enable SSH using /etc/init.d/SSH start, identify the IP address, SSH to the in-memory ESX host, and run vdq -q to list all storage devices.
In the example KS-ESX01.CFG, the ESX installation disk will be t10.NVMe____SKHynix_HFS512GEJ9X162N_________________AJEBN74041040C403___00000001 and NVMe Tiering disk will be t10.NVMe____MTFDLBA1T0THJ2D1BP15ABYY_________________5D19D3500175A000
After creating the bootable ESX installer on your USB device, copy your modified ESX kickstart file(s) into the root directory of the USB device (kickstart file names must be uppercase).
Now, navigate into the USB device under EFI/BOOT and edit BOOT.CFG updating the kernelopt so it matches the following which will run our KS-ESX01.CFG instead of the interactive installation:
bootstate=0
title=Loading ESXi installer
timeout=5
prefix=
kernel=/b.b00
kernelopt=ks=usb:/KS-ESX01.CFG
modules=/jumpstrt.gz --- /useropts.gz --- /features.gz --- /k.b00 --- /uc_intel.b00 --- /uc_amd.b00 --- /uc_hygon.b00 --- /vmx.v00 --- /vim.v00 --- /tpm.v00 --- /sb.v00 --- /s.v00 --- /atlantic.v00 --- /bcm_mpi3.v00 --- /bnxtnet.v00 --- /bnxtroce.v00 --- /cndi_igc.v00 --- /elxnet.v00 --- /i40en.v00 --- /iavmd.v00 --- /icen.v00 --- /igbn.v00 --- /intelgpi.v00 --- /ionic_cl.v00 --- /ionic_en.v00 --- /irdman.v00 --- /iser.v00 --- /ixgben.v00 --- /lpfc.v00 --- /lpnic.v00 --- /lsi_mr3.v00 --- /lsi_msgp.v00 --- /lsi_msgp.v01 --- /ne1000.v00 --- /nenic_en.v00 --- /nenic.v00 --- /nfnic.v00 --- /nhpsa.v00 --- /nipmi.v00 --- /nmlx5_cc.v00 --- /nmlx5_co.v00 --- /nmlx5_rd.v00 --- /ntg3.v00 --- /nvme_pci.v00 --- /nvmerdma.v00 --- /nvmetcp.v00 --- /nvmxnet3.v00 --- /nvmxnet3.v01 --- /pvscsi.v00 --- /qat.v00 --- /qcnic.v00 --- /qedentv.v00 --- /qedrntv.v00 --- /qfle3.v00 --- /qfle3f.v00 --- /qfle3i.v00 --- /rdmahl.v00 --- /rshim_ne.v00 --- /rshim.v00 --- /sfvmk.v00 --- /smartpqi.v00 --- /vmkata.v00 --- /vmksdhci.v00 --- /vmkusb.v00 --- /vmw_ahci.v00 --- /bmcal.v00 --- /clusters.v00 --- /crx.v00 --- /drivervm.v00 --- /btldr.v00 --- /dvfilter.v00 --- /esx_ui.v00 --- /esxupdt.v00 --- /tpmesxup.v00 --- /weaselin.v00 --- /xorg.v00 --- /esxio_co.v00 --- /infravis.v00 --- /loadesx.v00 --- /hpv2_hps.v00 --- /intelv2_.v00 --- /lsiv2_dr.v00 --- /nvme_pci.v01 --- /oem_dell.v00 --- /oem_leno.v00 --- /smartpqi.v01 --- /native_m.v00 --- /nsx_pyth.v01 --- /podvm_ro.v00 --- /qlnative.v00 --- /trx.v00 --- /vcls_pod.v00 --- /vdfs.v00 --- /vds_vsip.v00 --- /vmware_e.v00 --- /vmware_f.v00 --- /hbrsrv.v00 --- /vsan.v00 --- /vsanheal.v00 --- /vsanmgmt.v00 --- /tools.t00 --- /qp_esx_d.v00 --- /nsx_adf.v00 --- /nsx_cfga.v00 --- /nsx_cont.v00 --- /nsx_cpp_.v00 --- /nsx_esx_.v00 --- /nsx_expo.v00 --- /nsx_head.v00 --- /nsx_host.v00 --- /nsx_moni.v00 --- /nsx_mpa.v00 --- /nsx_nest.v00 --- /nsx_neto.v00 --- /nsx_opsa.v00 --- /nsx_plat.v00 --- /nsx_prot.v00 --- /nsx_prox.v00 --- /nsx_pyth.v00 --- /nsx_pyth.v02 --- /nsx_scx.v00 --- /nsx_sfhc.v00 --- /nsx_shar.v00 --- /nsx_snpr.v00 --- /nsxcli.v00 --- /vsipfwli.v00 --- /gc.v00 --- /imgdb.tgz --- /basemisc.tgz --- /resvibs.tgz --- /esxiodpt.tgz --- /imgpayld.tgz
build=9.1.0-0.25370933
updated=0
📒 If you are performing the installation on two or three physical ESX hosts, the only required change is to update the reference to kickstart file, so you know which file is being referenceed on the USB device.
-
Plug the USB device into your system and power on to begin the ESX installation.
-
Once ESX reboots (there is a secondary reboot as part of the ESX scripted installation), you should be able to log in to your ESX host using the FQDN and see something like the following:
- Deploy the VCF Installer appliance (VCF-SDDC-Manager-Appliance-9.1.0.0.25371088.ova) using the following shell script, deploy_vcf_installer.sh, which relies on OVFTool. Install OVFTool if you do not already have it on your local system.
- After the VCF Installer is up and running, we need to make a few configuration changes (credentials/networking). I have automated these changes, and you can run the following PowerShell script: setup_vcf_installer.ps1.
- Before we can deploy our VCF 9.1 environment, we need to connect to the VCF Offline Depot that you set up in Step 1.
Open your browser to the VCF Installer (e.g. https://sddcm01.vcf.lab/), log in with the password you configured in Step 8, and then click the DEPOT SETTINGS AND BINARY MANAGEMENT button.
If Step 8 successfully connected to your VCF Offline Depot, it should already have synced the VCF metadata, which shows the list of binaries available for download. Click the DOWNLOAD button to begin downloading the required VCF binaries and ensure each item listed in the table has a Success status.
- Navigate back to the VCF Installer homepage and click on
DEPLOY USING JSON SPECto begin your VCF deployment.
Upload your modified VCF deployment JSON and click Next to begin the validation.
- VCF 9.1 One Host Deployment JSON Samples
- VCF 9.1 Two Host Deployment JSON Samples
- VCF 9.1 Three Host Deployment JSON Samples
📒 You may encounter pre-checks that require acknowledgment to continue. I have noticed that with certain MikroTik devices, even when Jumbo Frames (MTU=9K) are configured, validation can fail while deployment still succeeds. In that case, simply acknowledge the configuration.
Once you have fixed and/or acknowledged all applicable pre-checks, click DEPLOY to start the deployment.
📒 During deployment, the VCF Installer generates its own self-signed TLS certificate, which forces a browser refresh and displays a message like the following. This is expected. Simply click the reload button as shown in the screenshot below.
- The deployment will take a few hours. Once everything has been deployed, you should see a success page like the following:
You can log in to your new VCF 9.1 deployment by clicking the link to VCF Operations and using the admin credentials you specified in your deployment JSON, or simply copying the credentials from this screen.
Here is an example of single host deployment and total deployment duration for each component:
- There are a number of post-VCF deployment optimizations that should be run for lab environments. You can refer to the linked blog post and run the PowerShell script
vcf-post-deploy-lab-tweaks.ps1.
- Additional optimization for those using AMD Zen4/5 CPUs.
- Apply additional NSX optimization for reducing High CPU Utilization on NSX components (NSX Manager, Edges and Virtual Network Appliances) when using AMD Zen4/5 CPUs.
- Apply additional VCFA optimization for reducing High CPU Utilization on VCVF Automation (VCFA) when using AMD Zen4/5 CPUs.