Cloud & DevOps Engineering (Part 6) : Understanding VMware Tanzu — From Physical Servers to Kubernetes

Bawantha Rathnayake

7 min read

·

Jul 17, 2026

--

Listen

Share

When I first started learning VMware Tanzu, one question kept coming to my mind:

“I understand Pods, Deployments, and Services… but where exactly does Kubernetes run?”

Most Kubernetes tutorials focus on Pods, Deployments, and Containers. However, in enterprise environments such as banks, telecom providers, and large organizations, Kubernetes doesn’t run directly on physical servers.

Instead, there are several layers working together:

• Physical Servers
• Virtualization
• Hypervisors
• VMware ESXi
• vCenter
• Supervisor Clusters
• Tanzu Kubernetes Clusters (TKG)

Understanding these layers is essential if you’re working with VMware Tanzu or enterprise Kubernetes platforms.

In this article, we’ll walk through the journey from physical hardware all the way to applications running inside Kubernetes Pods.

Press enter or click to view image in full size

Life Before Virtualization

Years ago, organizations commonly deployed one application on one physical server.

For example:

Payroll Application
        |
Physical Server 1

CRM Application
        |
Physical Server 2

Banking Application
        |
Physical Server 3

This approach worked, but it had major drawbacks:

• Expensive hardware costs
• Poor resource utilization
• Difficult maintenance
• Limited scalability

Many servers were using less than 20% of their available CPU and memory while still consuming power and rack space.

Organizations needed a better solution.

What is Virtualization?

Virtualization allows multiple virtual computers to run on a single physical server.

Press enter or click to view image in full size

Each Virtual Machine (VM) behaves like an independent computer.

Benefits include:

• Better resource utilization
• Reduced infrastructure costs
• Faster provisioning
• Easier scaling
• Improved disaster recovery

What is a Hypervisor?

A Hypervisor is software that creates and manages Virtual Machines.

It sits between physical hardware and virtual machines.

Virtual Machines
        |
    Hypervisor
        |
Physical Hardware

Hypervisor Types

There are two main types of hypervisors.

Press enter or click to view image in full size

Type 1 Hypervisor (Bare Metal)

A Type 1 Hypervisor runs directly on physical hardware.

Examples:

• VMware ESXi
• Microsoft Hyper-V
• Xen

These are commonly used in enterprise production environments.

Type 2 Hypervisor

A Type 2 Hypervisor runs on top of an operating system.

Examples:

• VirtualBox
• VMware Workstation

These are typically used for learning and development environments.

Understanding VMware ESXi

VMware ESXi is VMware’s Type 1 Hypervisor.

It is installed directly on physical servers and enables multiple virtual machines to share hardware resources.

Press enter or click to view image in full size

ESXi is responsible for:

• Creating virtual machines
• Managing compute resources
• Managing storage resources
• Managing networking

Each ESXi host can run dozens or even hundreds of virtual machines.

Managing Multiple ESXi Hosts

Imagine a bank has several physical servers:

Dell Server 1 → ESXi-1
Dell Server 2 → ESXi-2
Dell Server 3 → ESXi-3

Managing each host separately becomes difficult.

This is where vCenter comes into the picture.

What is vCenter?

vCenter is VMware’s centralized management platform.

Instead of logging into every ESXi host individually, administrators manage the entire environment through a single interface.

                vCenter
                    |
     --------------------------------
     |              |              |
   ESXi-1         ESXi-2         ESXi-3

Using vCenter, administrators can:

• Create virtual machines
• Monitor resources
• Configure storage
• Configure networking
• Perform VM migrations
• Enable High Availability (HA)

Think of vCenter as the control tower of the VMware infrastructure.

What is vSphere?

One common beginner question is:

“Is vSphere the same as ESXi?”

Not exactly.

ESXi is the hypervisor.

vCenter is the management platform.

vSphere is the complete VMware virtualization ecosystem.

It includes:

• vCenter
• ESXi Hosts
• Virtual machines(VMs)

Press enter or click to view image in full size

Enter Kubernetes

At this point, we have:

Physical Servers
      |
    ESXi
      |
 Virtual Machines

Traditionally, organizations would manually install Kubernetes on these virtual machines.

While this works, managing Kubernetes at enterprise scale becomes increasingly complex.

VMware introduced Tanzu to simplify Kubernetes operations inside vSphere.

What is VMware Tanzu?

VMware Tanzu is VMware’s Kubernetes platform.

It integrates Kubernetes directly into the vSphere ecosystem.

This allows infrastructure teams and platform teams to manage Kubernetes using familiar VMware tools.

Instead of treating Kubernetes as a separate platform, Tanzu makes Kubernetes a first-class citizen inside VMware infrastructure.

Understanding the VMware Tanzu Architecture Diagram

Press enter or click to view image in full size

Now that we’ve covered the individual concepts, let’s look at how everything works together in a real VMware Tanzu environment.

The following architecture shows the complete journey from physical hardware all the way to applications running inside Kubernetes Pods.

1. Physical Dell Servers

At the very bottom of the stack are the physical servers.

In this example, we have multiple Dell PowerEdge servers:

Dell Server 1
Dell Server 2
Dell Server 3
...
Dell Server N

These servers provide the actual hardware resources:

• CPU
• Memory (RAM)
• Storage
• Network Interfaces

Everything else in the architecture ultimately runs on these physical machines.

2. ESXi Hosts

Installed on each physical server is VMware ESXi.

Physical Server
       |
      ESXi

ESXi is VMware’s Type-1 Hypervisor.

Its job is to:

• Create Virtual Machines
• Allocate CPU resources
• Allocate Memory resources
• Manage Storage
• Manage Networking

Because ESXi runs directly on hardware, it provides high performance and is commonly used in enterprise production environments.

3. Supervisor Cluster

The next layer is the Supervisor Cluster.

ESXi Hosts
      |
Supervisor Cluster

Unlike a normal Kubernetes cluster that runs applications, the Supervisor Cluster is responsible for managing Kubernetes infrastructure.

Think of it as the Kubernetes management layer running on top of vSphere.

Its responsibilities include:

• Managing Kubernetes infrastructure
• Creating Tanzu Kubernetes Clusters
• Managing storage integration
• Managing networking integration
• Managing cluster lifecycle

Its key components include:

API Server

The entry point for Kubernetes operations.

Whenever users create resources such as:

• Namespaces
• Clusters
• Pods

the request is sent to the API Server.

Scheduler

The Scheduler decides where workloads should run.

It evaluates available resources and determines the best placement for workloads.

Controller Manager

The Controller Manager continuously monitors the desired state of the cluster and ensures the actual state matches it.

For example:

Desired Pods = 3
Current Pods = 2

The Controller Manager automatically creates the missing Pod.

Supervisor Control Plane VMs

The Supervisor Cluster itself runs on Virtual Machines.

In the diagram, these are shown as:

Supervisor Control Plane VM 1
Supervisor Control Plane VM 2
Supervisor Control Plane VM 3

These VMs are distributed across ESXi hosts for high availability

4. Supervisor Namespaces

Once the Supervisor Cluster is available, administrators create Namespaces.

Think of a Namespace as a dedicated workspace inside the Supervisor Cluster.

A Namespace helps with:

• Resource isolation
• Team separation
• Access control
• Storage allocation
• Network policies

Supervisor Cluster
        |
+------------------------+
| Namespace: Production  |
+------------------------+
+------------------------+
| Namespace: Development |
+------------------------+

Inside the Supervisor Cluster are Namespaces.

Think of a Namespace as a dedicated workspace.

The diagram shows examples such as:

Management Namespace
vSphere System Namespace
TKG System Namespace
Project Namespaces

5. Tanzu Kubernetes Cluster (TKG)

A Tanzu Kubernetes Cluster (TKG) is where applications actually run.

The Supervisor Cluster creates and manages TKG Clusters.

Each TKG Cluster contains:

    TKG Cluster
Control Plane Nodes
        |
  Worker Nodes
        |
      Pods
        |
   Applications

The Control Plane manages Kubernetes.

Worker Nodes run workloads.

In the diagram:

Production TKG Cluster

The TKG Cluster contains:

Control Plane VMs

These VMs run Kubernetes management components such as:

• API Server
• Scheduler
• Controller Manager
• etcd

The Control Plane manages the Kubernetes cluster.

Worker VMs

The Worker VMs are where application workloads run.

For example:

Worker VM 1
Worker VM 2
Worker VM 3

When Kubernetes schedules Pods, they are deployed onto these Worker Nodes.

6. Application Pods

Where Do Pods Run?

A common beginner misconception is:

“Do Pods run inside the Supervisor Cluster?”

No.

Pods run inside Worker Nodes that belong to a TKG Cluster.

TKG Worker Node VM
        |
       Pod
        |
    Container
        |
   Application

Applications such as:

• Spring Boot Applications
• React Applications

run inside Pods on Worker Nodes.

Inside the Worker Nodes, Kubernetes runs Pods.

The diagram shows three example workload types:

Web Front-End Pods

Application Pod 1
Application Pod 2
Application Pod 3

These could be:

• React applications
• Angular applications
• Web portals

Database Pods

Database Pod 1
Database Pod 2

These provide persistent data storage for applications.

Message Queue Pods

Message Pod 1
Message Pod 2

Examples include:

• RabbitMQ
• Kafka
• ActiveMQ

These enable communication between services.

Press enter or click to view image in full size