VMware Cloud Foundation Overview

VCF 9.0Beginnersupportadmin⏱ 60 min

VCFFTS9 course content — VCF 9.0 focused

Objectives

Explain VCF fleet architecture (Fleet → Instance → Management Domain → Workload Domains)
Identify and describe the eight VCF 9.0 use cases from the official curriculum
Differentiate between Simple and HA deployment models
List VCF core components and add-ons
Explain VCF Automation self-service catalog and IaC capabilities
Describe VCF Edge for edge computing use cases

Prerequisites

Active Holodeck VCF 9.0 lab or access to VCF documentation

Required skills:

Basic VMware terminology

Tasks

Task 1 VCF 9.0 Architecture & Core Components

Understanding VCF's layered architecture (Fleet → Instance → Domain) is foundational. A common confusion is conflating VCF Instance with vCenter instance — they are distinct constructs. VCF Operations manages the fleet; SDDC Manager manages the instance.

Understand VCF fleet architecture, core components, and deployment constructs

Step 1

VCF Private Cloud Constructs

VCF 9.0 introduces the fleet concept: VCF Fleet → VCF Instance → Management Domain + Workload Domains. A VCF Fleet includes VCF Operations, VCF Automation, and one or more VCF Instances. Each VCF Instance must have a management domain; workload domains are optional. VCF Operations provides the administrative console for the entire fleet. VCF Automation provides self-service IaaS and tenant management.

Step 2

VCF Core Components

VCF integrates: vSphere (compute), vSAN (storage), NSX (networking), VCF Operations (monitoring/fleet mgmt), VCF Automation (self-service/IaC). VCF Installer deploys the entire SDDC stack. SDDC Manager handles lifecycle management from bring-up through patching and upgrading. In VCF 9.0, VCF Operations, VCF Operations Fleet Management, and VCF Automation are now MANDATORY components (previously optional in VCF 5.x). This is a significant architectural change from earlier versions.

Step 3

VCF Deployment Models

Two deployment models via VCF Installer: Simple (single node for NSX Manager, VCF Operations, VCF Automation — not for production) and High Availability (three-node clusters with auto VIP for NSX Manager, preconfigured LB for VCF Operations, automatic LB for VCF Automation).

[HOLODECK NOTE] Holodeck deploys VCF in a nested environment on a single physical host. Only the Simple (single-node) model is practical in Holodeck. HA (3-node) requires significant RAM (384GB+ physical). Nested ESXi hosts share physical resources — performance characteristics differ substantially from bare-metal production deployments.

Step 4

VCF Add-Ons

VCF can be extended with add-ons: Live Recovery (ransomware + disaster recovery), vDefend (micro-segmentation, lateral security), Avi Load Balancer (software-defined LB with cloud console), VMware Private AI Foundation with NVIDIA, Tanzu (Kubernetes).

Validation Gate

Check: Can you draw the VCF hierarchy from Fleet down to Workload Domain and name the management component at each level?

Expected: Fleet (VCF Operations + VCF Automation) → Instance (SDDC Manager) → Management Domain (vCenter, NSX, vSAN) → Workload Domains (independent vCenter, NSX, vSAN per domain)

Common Errors

Confusing VCF Instance with vCenter instance

Fix: A VCF Instance = Management Domain + Workload Domains managed by one SDDC Manager. A vCenter instance is one component within a domain. One VCF Instance typically has 2-4 vCenter instances (one per domain).

Deploying Simple model in production

Fix: Simple deployment uses single-node NSX Manager, VCF Operations, and VCF Automation — no HA. This is for labs and PoCs only. Production requires HA deployment with 3-node clusters. VCF Installer explicitly warns against Simple for production.

Treating VCF Operations and VCF Automation as optional in VCF 9.0

Fix: In VCF 5.x, Aria Operations and Aria Automation were optional add-ons. In VCF 9.0, VCF Operations and VCF Automation are mandatory components deployed during bring-up. Designs that omit them are invalid.

Task 2 VCF 9.0 Use Cases — The Eight Scenarios

The eight use cases map directly to VCF component capabilities. Understanding which components enable which use cases is critical for both exam and real-world design — a customer asking for 'private AI' needs NVIDIA GPU passthrough + VCF, not just VCF standalone.

Identify and explain the eight primary VCF use cases from the official curriculum

Step 1

Building & Operating Private Cloud Infrastructure

VCF provides integrated private cloud platform deployable in hours not months. Key capabilities: integrated platform (compute/network/storage/mgmt/security), VCF Edge for edge use cases (smart manufacturing, IoT, retail, ROBO), automated deployment and lifecycle management via SDDC Manager, intelligent operations via VCF Operations, centralized fleet management, import of existing vSphere/vSAN environments.

Step 2

Implementing Automated Infrastructure to Deliver IaaS

VCF Automation delivers self-service IaaS. Key features: Self-service catalog (IaC templates → catalog items for low-tech users), IaC blueprints (visual canvas rendered in YAML), Terraform integration (HashiCorp provider), Orchestration and extensibility (VCF Operations orchestrator, event broker), Version control (GitHub/GitLab/BitBucket). Supervisor creates K8s control plane on hypervisor — enables VKS, VM Service, Network Service, Storage Services.

Step 3

Extending Data Center to Public Cloud

VCF Operations HCX enables workload migration from vSphere 6.x, live migration via HCX-assisted vMotion. License portability across on-prem, VCSP partners, and hyperscalers (AWS/GCVE/AVS). Consistent infrastructure and operations across environments. Bidirectional workload portability.

Step 4

Building, Running & Managing Modern Workloads

Single platform for VMs and containers. VKS = natively integrated CNCF-compliant Kubernetes runtime at no extra cost. VKS decoupled from vCenter — async K8s version updates. Supervisor enables provisioning IaaS services. VCF Automation services (formerly Supervisor Services): VKS, VM Service, Network Service, Storage Services, Object Storage, Image Registry (Harbor), Backup/Recovery, VCF Data Services.

Step 5

Private AI, Security & Disaster Recovery

Private AI: VMware Private AI Foundation with NVIDIA for GPU-accelerated workloads on VCF. Security: end-to-end protection with micro-segmentation, data encryption (at rest + in transit), RBAC, compliance checks, vDefend distributed lateral security. Ransomware Protection: Live Recovery single console for ransomware + DR recovery. Disaster Recovery: stretched clusters, vSphere HA, vSAN Data Protection.

Validation Gate

Check: List the eight VCF use cases and identify which require add-on licensing

Expected: Base VCF covers: private cloud, IaaS automation, cloud extension (HCX), modern workloads (VKS). Add-ons required: vDefend (security), Live Recovery (DR/ransomware), Avi ALB (load balancing), Private AI (NVIDIA GPU)

Common Errors

Conflating VKS with standalone Tanzu Kubernetes Grid

Fix: VKS (VMware Kubernetes Service) is natively integrated into VCF via Supervisor. It is NOT the same as TKG standalone. VKS is decoupled from vCenter for async K8s updates and is included at no extra cost with VCF.

Assuming Live Recovery is included in base VCF

Fix: Live Recovery (ransomware + DR) is a paid add-on, not part of base VCF subscription. Same for vDefend, Avi Load Balancer, and Private AI Foundation. Only VCF Operations and VCF Automation are mandatory/included.

Design Reflection (VCDX)

VCF architecture questions are foundational in VCDX defense. Panelists expect you to explain why you chose VCF over à la carte vSphere + NSX, and to articulate the operational value of the fleet model for multi-instance management.

Requirements

Understand VCF fleet hierarchy and component relationships
Differentiate Simple vs HA deployment models
Map use cases to required components and add-ons

Constraints

VCF 9.0 mandates VCF Operations + Automation — cannot opt out
Simple deployment model not supported for production

Assumptions

Students have basic VMware terminology knowledge
Lab environment uses Holodeck (Simple deployment)

Risks

Designing for production using Simple model parameters observed in Holodeck lab
Omitting mandatory components from architecture diagrams

Self-Assessment Discussion Prompts

Why did VMware make VCF Operations and VCF Automation mandatory in VCF 9.0 when they were optional in 5.x?
When would you recommend separate VCF Instances vs a single Instance with multiple workload domains?
How does the VCF Fleet model change the operational model compared to managing individual vCenter instances?

⚠ Known Pitfalls (from Community KB)

Studying VCF 5.x architecture and assuming it applies to 9.0 — the Fleet/Instance/Domain hierarchy, mandatory components, and naming conventions changed significantly.

Ignoring add-on vs included licensing boundaries — this is a common exam trap and a real-world budgeting error.

References

VCF 9.0 Architecture OverviewTier 1 — Official
VCF 9.0 Planning and Preparation GuideTier 1 — Official

Mark complete