Guides / Walkthroughs

Step-by-step walkthroughs

Longer than a post, maintained as the products move. Each one has been run end to end in the lab — personal use, so treat with appropriate caution: your environment will differ.

Filter:

VCF 9.1.1 Lab

12 guides
expand_more
01.

VCF 9.1.1 Lab Rebuild

The starting page for the VCF 9.1.1 lab rebuild. What the guide is, how it is written, how to read it, and the list of pages as they are added.

vcf vcf911 lab rebuild
arrow_forward
02.

The Hardware

Eight machines in two clusters: four identical AMD mini-workstations for the management cluster on vSAN ESA, and four different Intel machines for workloads on NFS. Fleet totals, memory per host with NVMe tiering, a table per cluster, and the five things the hardware decides before the build starts — including where the lab departs from the docs.

vcf vcf911 hardware esx vsan memory-tiering lab
arrow_forward
03.

Kickstart: One USB Key, Eight Hosts

The ESX install for the lab. The kickstart is rendered from the host inventory, not written by hand, and a check fails the build if anyone edits it. The root password is a pointer until the moment the USB key is written. Covers writing the key, the unattended install, the kickstart line by line against the VCF host preparation requirements, and the lab-only settings for consumer AMD hardware.

vcf vcf911 esx kickstart automation secrets lab
arrow_forward
04.

Networks, Names and Addresses

The reference page for the rebuild. What the external VLAN serves and why everything above it is dynamic while the VLAN itself is not. Five VLANs for bring-up and why the management network is a VLAN and not an overlay, what the distributed gateway's external VLAN demands of the physical network, three DNS zones, sixteen component names, the two node ranges that must hold no names, and how all of it was proven before the Installer was opened.

vcf vcf911 networking vlan dns ip-addressing nsx vpc lab
arrow_forward
05.

The VCF Installer: Drive It from JSON, Check It in the Wizard

The VCF Installer for the 9.1.1 rebuild. The deployment is driven from a JSON file and the wizard is used to check it, and the build found a reason for each half of that rule. Covers the depot and binaries, the Plan choices for Full Stack VPC with Distributed Connectivity and vSAN ESA, the Installer's own JSON template, a first validation run that failed on one check and skipped fifteen, a clean second run on lab hardware, and the distributed gateway network that validation does not test.

vcf vcf911 installer deployment json nsx vpc vsan-esa lab
arrow_forward
06.

Bring-Up: Five and a Half Hours, 320 Subtasks

The VCF 9.1.1 bring-up from Deploy to a complete instance: 2026-09-17 23:00 to 04:21, eight milestones, 320 subtasks, none failed. Why subtask counts predict nothing about duration, the connectivity error that is not a failure, how to watch progress without the browser session, what the instance could and could not do when it finished, and what the Installer left behind for the credential store.

vcf vcf911 deployment bring-up sddc-manager nsx lab
arrow_forward
07.

Distributed Connectivity: What the Physical Network Owes the Instance

The part of distributed connectivity that lives on the gateway and the switch. Why this model uses no Edge cluster and no BGP, and why the distributed VXLAN model does. Six requirements taken from the documentation, each with a check that produces evidence: the VLAN and its gateway, ICMP redirects across all three sysctl scopes, the hairpin, zoning, and proving the external VLAN is trunked to the TEP uplinks without opening SSH on a host. Plus the Network span field and what selecting it does.

vcf vcf911 nsx vpc networking distributed-connectivity validation lab
arrow_forward
08.

The VNA Cluster: Where Stateful Services Live When There Is No Edge

The distributed transit gateway is instantiated in every host's datapath, which is what lets traffic leave the host directly and also means there is nowhere for a connection table to live. So stateful services move to a VNA cluster. The dependencies, the two-node deployment at a size below the recommendation, a recommendation that reads like it asks for a new VLAN and does not, two API details, and a throwaway VPC used to measure the data path end to end.

vcf vcf911 nsx vpc vna nat networking lab
arrow_forward
09.

The Second Cluster, As Built

The second cluster as it exists on 2026-09-19, every figure read from vCenter after the build: hosts, EVC, the distributed switch, VMkernel addressing, six NFS datastores with their tags and folder, and how it was built. It was built in vCenter and is not managed by SDDC Manager. The page ends with what that limits, why SDDC Manager's add-cluster validators would not run on these hosts, and the plan to rebuild it through SDDC Manager.

vcf vcf911 vsphere cluster nfs evc distributed-switch lab
arrow_forward
10.

Attaching the Second Cluster to NSX

NSX host preparation for pgvcf1-cl02, done from the cluster's own Network Configuration page in vCenter. The five inputs, which of them differ from the dialog's defaults and why, what NSX on DVPG does, what preparation created, a second tunnel endpoint per host on an uplink with no NIC behind it, and packet loss from one host traced to a switch-port setting.

vcf vcf911 nsx vpc tep transport-nodes lab
arrow_forward
11.

Avi Load Balancer: One Controller, NSX Cloud in VPC Mode

Avi goes in before the Supervisor, because the Supervisor's VPC wizard has no load balancer field — NSX picks Avi only if it is already registered. The values used and why, the seven steps from the Service Engine network to a VPC enabled for Avi, what to see at each one, and the result: virtual services with VIPs allocated by NSX out of the external IP block, on a distributed transit gateway with no Edge cluster.

vcf vcf911 avi load-balancer nsx vpc lab
arrow_forward
12.

Two Supervisors on VPC Networking

Activating a Supervisor on VPC networking with a distributed transit gateway and Avi — no Edge cluster, no Tier-0, no BGP. The values that can never be changed and are therefore decided first, storage policies whose names become StorageClass names, the seven wizard steps with what each one takes, what happens in the 25 minutes after Finish, and the second Supervisor activated from a configuration file.

vcf vcf911 supervisor kubernetes vks nsx vpc avi lab
arrow_forward

VCF 9.1 Lab

4 guides
expand_more

VCF 9.0.2 Lab

15 guides
expand_more
01.

Lab Environment Overview

At-a-glance reference for the entire pgnet.io lab: 3-node consolidated VCF 9.0.2 cluster on AMD Ryzen hardware, 11-VLAN network schedule, full service inventory with FQDNs and IPs, DNS/identity strategy with BIND 9 and Active Directory, and links to every detailed guide.

vcf lab topology networking reference
arrow_forward
02.

Architecture & Planning

Pre-deployment planning for a 3-node consolidated VCF 9.0.2 cluster. Covers hardware BOM, lab-specific workarounds (vSAN ESA Mock VIB, AMD Ryzen), VLAN schedule across 11 networks, full DNS validation table, and identity strategy using BIND 9 with AD forwarding.

vcf architecture planning
arrow_forward
03.

Infrastructure Preparation

Physical switch config for jumbo frames and VLAN trunking, then automated ESXi bootstrapping via Kickstart — disk wipe, OS install, vmk0 networking, SSH hardening, Memory Tiering, vSAN ESA Mock VIB, and AMD Ryzen workarounds across all three hosts.

vcf infrastructure esxi networking
arrow_forward
04.

Deployment (Bring-Up)

Deploy the SDDC Manager OVA then run the VCF bring-up wizard: depot connection, deployment spec (hosts, vSAN ESA, NSX, network profiles), full JSON config reference, and troubleshooting commands for DNS, NTP, and MTU validation.

vcf deployment sddc-manager
arrow_forward
05.

Post-Deployment: Identity & SSO

Configure the AD identity source over LDAPS, create the embedded identity broker SSO domain, map AD groups to VCF roles, and assign administrator access across vCenter, NSX Manager, and VCF Operations.

vcf identity sso active-directory ldap
arrow_forward
06.

Post-Deployment: Certificates

Register the Microsoft CA in Fleet Manager, generate CSRs for SDDC Manager, vCenter, and NSX, submit to the CA, import signed certs, and run the Replace Certificates workflow to propagate trust.

vcf certificates pki tls
arrow_forward
07.

Post-Deployment: Networking (BGP & NSX Edge)

Configure FRR BGP on the physical router, deploy two NSX Edge nodes via the VPC wizard, set TEP pool, configure the Tier-0 Gateway with BGP ASN 65001, and wire up dual uplinks on both pgen1 and pgen2.

vcf nsx networking bgp edge
arrow_forward
08.

Post-Deployment: Supervisor Deployment

Deploy the Supervisor cluster from vCenter using VCF Networking, configure management network, vSAN storage policy, NSX VPC settings, and validate the Supervisor API endpoint — prerequisite for all VKS workloads.

vcf supervisor kubernetes vks nsx
arrow_forward
09.

Post-Deployment: Lab Features

Lab-only extras — enable root SSH on SDDC Manager via the vcf.sddc-root Salt state, and register a secondary vSphere cluster as a lifecycle deployment target to free up VCF domain resources.

vcf lab salt lifecycle
arrow_forward
10.

Operations

Early operations notes: alert policy configuration, suppression of expected non-redundancy warnings, and placeholder sections for ongoing alerting as the environment matures.

vcf operations
arrow_forward
11.

Logs Deployment & Use

Deploy VCF Operations for Logs from SDDC Manager, configure DNS/NTP/identity prerequisites, set up SSO, then configure log forwarding for vCenter, ESXi, NSX Manager, and SDDC Manager.

vcf logging observability
arrow_forward
12.

Automation

Initial VCF Automation setup: SSO via embedded Identity Broker, certificate standardisation, organisation and project config, capacity and networking quick start, and connectivity validation. Tenancy and VPC provisioning noted as ongoing.

vcf automation vpc
arrow_forward
13.

NSX & Network Operations

NSX Day 2 ops and VCF Operations for Networks onboarding — platform and collector appliance connection, identity config, BGP peering validation, and collector health checks. Requires BGP and NSX healthy from Part 4.

vcf nsx networking bgp
arrow_forward
14.

VKS & Supervisor Services

Deploy the full Supervisor services stack: VCF CLI context, Local Consumption Interface, Contour ingress, CA Cluster Issuer via cert-manager, Harbor with CA-issued TLS, and Argo CD operator — with ordered CLI and vCenter steps.

vcf kubernetes vks tanzu
arrow_forward
15.

Appendices

Reference appendices: FRR BGP config for the UDM router, full ESXi Kickstart script with all lab workarounds, SDDC Manager bring-up JSON, and an appliance credentials reference for all VCF component FQDNs.

vcf reference
arrow_forward

Individual Guides