Addresses pending for ever, a retryable error that never stops retrying, and an ordering rule the docs don’t tell you: in a self-service NSX VPC, the load balancer must exist before the namespace that uses it.
Three nested-ESXi pods with byte-identical addressing, down to the MACs, and no route between them. How overlapping VPC address ranges turn cookie-cutter environments into a feature.
Three pods with identical addressing all need the same repo, WSUS and AD. One shared-services VPC serves them all through the transit gateway, and can’t reach back into any of them.
The whole isolated pod, nested ESXi hosts and all, as one VCF Automation blueprint in the catalog. How the blueprint is built, the order it enforces, and the three things it can’t express.
A second vNIC on a nested host adds no physical redundancy, so why add it? We expected bringup to want one, and pulling vmnic0 mid-SSH proved the trunk copes: 0% loss, session intact.
The nested ESXi appliance sat at ‘waiting for DHCP’ for ever. VPC subnets don’t hand out addresses: the VM Service does, through cloud-init, sysprep or OVF guestinfo, depending on the guest.
Seven ways the nested-ESXi pod blueprint failed validation before it worked, from ${input} inside flow mappings to an image-sync race. Short, specific, and each one cost me a cycle.
VCF Automation 9.1 has two provisioning models, as two kinds of organisation. I ran the same use cases through both, from Windows VMs to nested ESXi. The honest scorecard, and how each case is done.
The same VKS cluster built twice: once with kubectl, once as a VCF Automation request. The Cluster object is identical. Everything around it isn’t, and the difference is what you get for free.
How an interactive PowerShell script grew into a catalog item that builds complete nested VCF 9.1 instances from one request form. The design rules that made it survivable, and the traps behind them.