Work top to bottom. Every field says what it is; where a field says Not sure? , that value is a safe starting point you can refine later with the VCF architect.
VCF fleet Platform
VMware Cloud Foundation (VCF) VMware vSphere Foundation (VVF) / standalone VCF follows the workbook's management domain (SDDC Manager, NSX, workload domains). VVF has none of those: vCenter, VCF Operations, a License Server and, optionally, the VCF Management Services. In VVF the management VMs usually share the cluster with your workloads.
vSphere Foundation profile not in the workbook
Built from the TechDocs 9.1 VVF deployment pages, with appliance sizes from the workbook tables. The VCF Installer deploys VVF with a Medium vCenter, a Medium VCF Operations node and Small management services; a manual build can leave the management services out. What it counts and why
VCF Management Services
Include (VCF Installer deployment) Leave out (manual deployment only) The services runtime with fleet and SDDC lifecycle, the software depot and telemetry. Not sure? Include.
Leaving the management services out is not offered by the VCF Installer : it deploys VVF only with them. You deploy vCenter, VCF Operations and the License Server by hand instead (guide ). There is then no Log Management and no software depot; the services can be added later with VCF Installer 9.1.1 or later.
The two fields below size the VCF Management Services (services runtime) only. The Installer uses Simple / Small ; a JSON spec can change it.
Deployment model
Simple High Availability Simple runs one copy of each management appliance — fine for labs and small sites, but a failure takes it offline. High Availability runs clustered copies that survive a failure. Not sure? High Availability.
Deployment size
Small Medium Large How large the environment is that VCF will manage. As a guide, the management vCenter at each size handles: Small up to 100 hosts / 1,000 VMs, Medium up to 400 hosts / 4,000 VMs, Large up to 1,000 hosts / 10,000 VMs. Simple is always Small. Not sure? Medium.
Instance model
First Instance Additional Instance First Instance : the first (or only) VCF instance — it also hosts the fleet-wide services. Additional Instance : a second site or region joining an existing VCF fleet. Not sure? First Instance.
Storage for the management cluster
vSAN-ESA vSAN-OSA NFS FC vSAN-ESA : storage built from the hosts' own NVMe disks (current generation). vSAN-OSA : the older vSAN with separate cache and capacity disks. NFS / FC : an external storage array. Not sure? vSAN-ESA.
The management cluster you intend to build — the tool checks whether everything above and below fits on it, even with one host down.
Cluster type
Standard Stretched (multi-AZ) Standard : one site. Stretched : the cluster spans two sites so it survives losing a whole site (at least 8 hosts, half per site). Not sure? Standard.
The stretched ×2 mirror is vSAN only . With NFS / FC stretched (vSphere Metro Storage Cluster), replication between sites is done by the storage array and is not modelled here — enter the datastore capacity available to the management cluster.
Number of hosts
CPU cores per host
Physical cores of all CPUs in one host together — not threads / hyperthreading.
CPUs (sockets) per host
Only used for the licence count: every CPU is licensed at 16 cores minimum .
RAM per host (GB)
Usable datastore capacity for the management cluster (GB)
CPU oversubscription (X:1)
How many virtual CPUs may share one physical core. Not sure? 1 (no overcommit — the workbook default; do not go above 2 for management).
RAM oversubscription (X:1)
How far memory may be overcommitted. Not sure? 1.
vSAN rebuild + operations reserve (%) vSAN only Free vSAN space kept for rebuilding after a host failure and for day-to-day operations. Not sure? 30.
Not applied to NFS / FC — only the storage growth % is added on top of VM disk + swap.
Storage growth (%)
Extra storage headroom for future growth. Not sure? 10.
Management infrastructure Add-ons
Optional infrastructure that runs in the management cluster. Leave anything you have not heard about in the design on Excluded .
Optional extras that most VVF environments do not use. Leave them on Excluded unless the design includes Kubernetes (Supervisor) or the Avi load balancer.
NSX Edges or Virtual Network Appliances
Excluded NSX Edge Small NSX Edge Medium NSX Edge Large NSX Edge XLarge VNA Small VNA Medium VNA Large VNA XLarge Route traffic between VCF's virtual networks (NSX) and the physical network. NSX Edge is the classic option, VNA the newer VPC option. Not sure? Excluded, unless the network design puts NSX routing on the management cluster.
NSX Global Manager
Excluded Small Medium Large XLarge Only for NSX Federation — one network and security policy spanning several sites. Not sure? Excluded.
vSphere Supervisor (Kubernetes) on the management cluster
Excluded Single Node High Availability Supervisor size
Tiny Small Medium Large Xlarge The Supervisor runs Kubernetes on vSphere. Only when Kubernetes runs in the management domain ; counts its 1 or 3 control-plane VMs. An Edge-backed Supervisor also needs the NSX Edges at Large or bigger. Not sure? Excluded.
The Supervisor runs Kubernetes on vSphere. Without NSX it uses vDS networking with a load balancer (Foundation Load Balancer or Avi). Counts its 1 or 3 control-plane VMs only. Not sure? Excluded.
Avi Load Balancer
Excluded Small Large X-Large VMware's load balancer (formerly NSX Advanced Load Balancer), e.g. for Kubernetes or VCF Automation. Counts its 3 controllers; the service engines run in the workload domains next to the workloads they serve and are not counted. Not sure? Excluded.
Security Services Platform (vDefend)
Excluded Include The platform behind vDefend advanced security (intrusion detection, malware prevention). Only with a vDefend licence. Not sure? Excluded.
vDefend / Avi licensing)A small licensing appliance. Needed for vDefend; for Avi only when it is licensed on-premises (Avi can also license directly from the cloud). Not sure? Tick it only if vDefend is bought.
A small licensing appliance, for Avi only when it is licensed on-premises (Avi can also license directly from the cloud). It is not the License Server, which is always counted. Not sure? Leave off.
vCenter size
Tiny Small Medium Large XLarge Small up to 100 hosts / 1,000 VMs, Medium up to 400 hosts / 4,000 VMs, Large up to 1,000 hosts / 10,000 VMs. Not sure? Medium (the Installer default).
vCenter storage
Default Large XLarge VCF Operations
Single Node HA - 2 nodes HA - 3 nodes VVF includes VCF Operations; you need it to license the hosts. Single Node is what the Installer deploys. HA - 2 nodes adds a replica; HA - 3 nodes adds a data node as well. Not sure? Single Node.
VCF Operations node size
Extra Small Small Medium Large Extra Large Use the Ops & Logs sizer or Broadcom's estimator for the right size. Not sure? Medium (the Installer default).
Cloud Proxy
Exclude Small Standard TechDocs lists the Cloud Proxy as optional for VVF. Not sure? Small.
The License Server (2 vCPU / 4 GB / 12 GB) is always counted: every VCF Operations instance that manages licences needs one.
VCF Operations
Include Existing Exclude The central management and monitoring console of VCF — every VCF fleet has one. Include : deploy it here (it brings a Cloud Proxy and, on a first instance, the License Server). Existing : it already runs elsewhere, e.g. when adding to an existing fleet — only its Cloud Proxy and the License Server are counted. Not sure? Include.
A Cloud Proxy collects data for VCF Operations. It is added automatically with VCF Operations; tick this only when VCF Operations is Excluded but a proxy should still run here.
Self-service portal for provisioning VMs and Kubernetes (formerly Aria Automation) — the private-cloud layer. Not sure? Leave off unless self-service is in scope.
VCF Automation size
Small Medium Large Small = 1 node, Medium / Large = 3 nodes. Not sure? The same as the deployment size above. VCF Operations for Networks
Excluded Small Medium Large XL XXL Network monitoring and traffic-flow analysis (formerly Aria Operations for Networks / vRNI). Not sure? Excluded.
These run inside VCF's own management platform (the services runtime ), so they do not add separate appliances — they add capacity to it. The result below shows how many runtime nodes that comes to.
Log Management size
Exclude Small Medium Large Central log collection and search (formerly Aria Operations for Logs). Not sure? Exclude, or Small for a small fleet.
Log Management replicas
Copies for capacity and availability — at least Small 1 / Medium 3 / Large 6, at most 19. Not sure? The minimum. Log Management requires the VCF Management Services . Set them to Include in section 1 to size it.
High-frequency (real-time) performance data in VCF Operations. Not sure? Leave off.
Only on an additional instance : a local copy of the VCF software repository, and a local sign-in (single sign-on) service. The first instance already has both.
Site Protection & Disaster Recovery
Exclude Management Only Workload Only Management & Workload Replication and failover to a second site with VMware Live Recovery (formerly Site Recovery Manager / vSphere Replication). Choose what is protected; mark protected workload domains below. Not sure? Exclude.
An isolated recovery environment for ransomware incidents. Only when that add-on is planned.
+ Add workload domain A workload domain is a separate vCenter + NSX environment where the customer's own workloads run. Its management VMs (vCenter, NSX Managers, and any Avi / vDefend controllers) live in this management cluster, so add one row per planned workload domain. NSX model : Shared reuses an existing NSX, Dedicated gets its own (HA Cluster = 3 managers).
experts Leave this off unless a design document specifies sizes that differ from the profile. It is the workbook's Advanced Management Domain Sizing .
Management vCenter size
Tiny Small Medium Large XLarge Management vCenter storage
Default Large XLarge NSX Manager model
HA Cluster Single Node NSX Manager size
Medium Large XLarge VCF Operations model
Exclude HA Cluster Single Node VCF Operations size
Exclude Extra Small Small Medium Large Extra Large Cloud Proxy size
Exclude Small Standard VCF Automation size
Exclude Small Medium Large While this is on, VCF Operations, its Cloud Proxy and VCF Automation follow these choices (Exclude removes them); the NSX Manager model also sets the host minimum (4 for an HA cluster).
