Lab Configuration

So the first part of this series is to quickly cover the lab environment. Now the majority of people who are seasoned lab users will probably skip this as all the information is available in many other already published series. But those who are new or following this to create their own environments for the first time might find the information useful.

As previously mentioned I was rushing to create and document the environment and I missed a number of key configuration items that caused the deployments to fail, costing me time. The biggest issue I had was the use of snapshots and remembering what these included and did not when I had to roll back, the amount of times I made a change, forgot to document it and when I reverted caused me the same issue again was frustrating but it certainly blew the cobwebs away and got me back into the labs with a bang. Once you have done this a couple of times it all becomes second nature especially when you remember to document it all and put proper notes in the snaps.

Before you start I would advise you to read the following blog in detail https://williamlam.com/2025/06/ultimate-lab-resource-for-vcf-9-0.html

Update : Originally when I planned on doing the demo I was going to also show vSAN ESA as well. I used the following blog https://williamlam.com/2025/02/vsan-esa-hardware-mock-vib-for-physical-esxi-deployment-for-vmware-cloud-foundation-vcf.html to install the VIB that would overcome the nested issues and the HCL. Now this worked perfectly and all the VVF validations worked and deployment started. However it came to a point in the workflow where it was creating the cluster image for vLCM and it failed extracting the VIB. The error was A general system error occurred: Failed to extract image from the host: an error occurred while re-creating VIB williamlam.com_bootbank_nested-vsan-esa-mock-hw_1.0.0-1.0. I tried several work arounds and could not get this to work, but strangely the same hosts and configuration worked on a VCF 9.0 deployment. Once this series is complete I will take the time to properly troubleshoot and update the blog on the ESA settings so you will have OSA and ESA options.

Documentation

If like me you like to try things out regularly in the lab, or like me miss things that mean you have to do multiple rebuilds until things are completely correct then documenting all the details will save you no end of time. A document with FQDNs, full IP details, NTP and passwords will allow you to quickly copy and paste into the install workflows and save you having to dig around or type it manually.

DNS

must have all entries for all components with forward and reverse lookup in place (including VCF Installer) all entries must be in lower case. This is important for certificate generation and resolution but is best practise as a VCF deployment will fail if you use upper case or mix

NTP

Access to working NTP is critical, this will be tested in he validation workflows and cause failures if it is out or not working

Nested ESX host configuration

OSA

cpu with present virtualisation or Error: Task failed on server: This host does not support Intel VT-x

ESA

To come at a later date https://williamlam.com/2025/02/vsan-esa-hardware-mock-vib-for-physical-esxi-deployment-for-vmware-cloud-foundation-vcf.html esxcli software acceptance set –level=CommunitySupported esxcli software vib install -d /tmp/nested-vsan-esa-mock-hw-offline-bundle.zip

Install ESX on the nested hosts

I wont be covering this in the blog as there are a million other blogs covering this in detail You can get the media using the following article https://knowledge.broadcom.com/external/article/401497/vmware-vcf-or-vvf-90-downloads-in-the-br.html We are using old kit in the lab so needed to use allowLegacyCPU=true to allow us to install ESX on the nested hosts (steps detailed in Williams blog) Once the install is complete go to the DCUI and configure the host with their IPs, names, DNS settings and name search Set SSH to be enabled

ESX Configuration

I know you are itching to get the VCF Installer deployed but there are still steps required on the hosts to make sure they pass all the validations that are built into the VVF deployment workflow.

You must do this on all the ESX nodes you will be using.

Using the remote console log onto the ESX server UI

You need to configure an NTP server or servers, set the service to start with the host and then start the service

Check that the server has the correct host name and FQDN in the settings, when you install an ESX server the server it will use default information when it creates its SSL certificate. Now you have changed the name to correct details you will need to re generate the certificate. If you do not do this step the VVF validation will fail and will not continue until you have rectified it SSH to the server run /sbin/generate-certificates Reboot the server

Do not change any other configuration like adding nics to switches, the VVF deployment process will take care of this and will fail validation if there is more than one nic is the default vSwitch

I recommend you take a snap of each ESX server at this point. This will enable you to quickly roll back if you have deployment issues.

Please note I have always had issue when creating nested ESX server snap images. The snap process works fine but in my experience it always has issues with the NTP configuration. If you do have to revert to a snap it would be work checking that the service is running and configured to start with the host before trying to run any steps again.