VMware ESXi and vSphere Cluster Management

What Is VMware ESXi? Type 1 Hypervisor Architecture and Core Concepts

Learn what VMware ESXi is, how its Type 1 hypervisor and VMkernel work, how it compares with Type 2 virtualization, and how ESXi fits into vSphere.

VMware ESXi is VMware's bare-metal, or Type 1, hypervisor. It installs directly on compatible physical server hardware and provides the virtualization layer on which virtual machines run. ESXi is the foundational compute component of the VMware vSphere ecosystem.

Instead of installing ESXi on top of Windows or Linux, an administrator installs it on the server itself. ESXi then presents virtual processors, memory, storage, and network devices to virtual machines while coordinating their access to the underlying physical resources.

What a Hypervisor Does

A hypervisor is software that creates, runs, isolates, and manages virtual machines. A virtual machine (VM) is a software-defined computer with virtual hardware. It can contain its own guest operating system and applications.

Without virtualization, one physical server commonly runs one primary operating system and its workloads. With ESXi, multiple VMs can share the same server. Each VM has its own virtual CPU, memory allocation, virtual disks, and network interfaces. The workloads remain logically separated even though they use shared physical hardware.

TermMeaning
Physical hostThe server containing the CPU, RAM, storage devices, network adapters, and firmware on which ESXi is installed.
HypervisorThe virtualization software that controls and schedules access to physical resources.
Virtual machineA software-defined computer presented to an operating system.
Guest operating systemThe operating system installed inside a VM.
ApplicationA workload running inside the guest operating system, such as a file service or web application.

Example: consolidating small workloads

A server running ESXi might host three separate VMs: one for a directory service, one for file storage, and one for an internal web application. Each VM has its own guest operating system and assigned virtual resources. The workloads share the server's physical CPU, RAM, storage, and network connections while remaining separate from one another.

Type 1 Versus Type 2 Hypervisors

A Type 1 hypervisor, also called a bare-metal hypervisor, runs directly on physical hardware. ESXi is a Type 1 hypervisor and is commonly used on production servers that run always-on business workloads.

A Type 2 hypervisor, also called a hosted hypervisor, runs as an application inside a conventional host operating system. Products such as VMware Workstation and Oracle VirtualBox are Type 2 examples. The hosted design depends on the host operating system for many hardware, I/O, and memory-management services.

CharacteristicType 1 / Bare-Metal HypervisorType 2 / Hosted Hypervisor
Where it runsDirectly on physical server hardware.Within a conventional host operating system.
Dependency on a host operating systemDoes not require a general-purpose host OS beneath the hypervisor.Relies on the host OS for many hardware and I/O services.
Typical use caseProduction servers, data centers, and infrastructure workloads.Desktop testing, development, training labs, and occasional workstation use.
Example productsVMware ESXi.VMware Workstation and Oracle VirtualBox.
Hardware and I/O pathThe hypervisor coordinates VM access to hardware.Requests pass through the guest, hosted hypervisor, host OS, and hardware.

The distinction is about architecture, not simply performance. A laptop running VirtualBox can be very useful for learning, while ESXi is designed around dedicated server hardware and centralized infrastructure operations.

ESXi Architecture and the VMkernel

The VMkernel is the ESXi hypervisor kernel. It receives resource requests from virtual machines, schedules and coordinates those requests, and manages access to the physical host's processors, memory, storage paths, and network interfaces.

The virtualization layer abstracts physical resources into virtual resources. A VM normally interacts with virtual hardware rather than directly controlling a physical disk, network adapter, or CPU core. This abstraction allows several VMs to use one host and lets ESXi apply isolation, scheduling, and resource controls.

The high-level request path looks like this:

  1. An application requests a service, such as reading data from a disk.
  2. The guest operating system sends the request to a device presented by the VM, such as a virtual storage controller.
  3. The VMkernel interprets, maps, and schedules the request through the appropriate virtual and physical storage paths.
  4. The physical hardware performs the operation and returns the result through the VMkernel and virtual device to the guest.
ComponentRoleExamples of Resources or Functions
Physical host hardwareProvides the actual compute, memory, storage, and network capacity.Processors, RAM, disks, storage controllers, and network adapters.
ESXi VMkernelCoordinates VM access to physical resources.CPU scheduling, memory management, storage paths, and network processing.
Virtual hardwarePresents standardized devices to a VM.vCPUs, virtual memory, virtual disks, and virtual network adapters.
Guest operating systemUses the VM's virtual hardware as if it were a computer.Windows, Linux, or another supported guest OS.
Application workloadRuns inside the guest OS.Databases, directory services, file services, and web applications.

How ESXi Virtualizes Host Resources

CPU virtualization

A vCPU is a virtual processor assigned to a VM. ESXi schedules vCPU work onto the host's physical processor execution resources. A logical CPU is a processor execution thread presented by the physical host to ESXi. Assigning more vCPUs does not create more physical processing capacity; it creates more scheduled work that must fit within the host's finite CPU capacity.

Memory virtualization

ESXi allocates host RAM to VMs and maintains mappings between guest memory addresses and physical memory. The guest OS sees virtual machine memory, while the VMkernel manages the underlying physical pages and resource sharing. Memory allocation must account for the needs of every VM and the host's own operations.

For deeper study, see Memory Virtualization Explained and Memory Overcommitment Explained.

Storage virtualization

VM files are stored in a datastore, which is storage presented to ESXi for virtual machine files and related data. A datastore can be backed by supported local or shared storage. The VM sees virtual disks, while ESXi maps those disks to datastore files and physical storage paths.

See Create a VMFS Datastore for a related storage concept.

Networking virtualization

A virtual switch, or vSwitch, is a software-based networking construct. It connects VM virtual network adapters to one another and to physical network uplinks. VLAN configuration, uplink selection, and physical switch configuration determine how VM traffic reaches other systems.

See Create Standard Switches and Configure VLANs for related topics.

Supported Hardware and Guest Operating Systems

ESXi requires compatible 64-bit, server-class hardware and processor virtualization support. Compatibility is specific to the combination of ESXi release, server model, processor, storage controller, network adapter, firmware, and installed drivers.

Before installation, check the vendor compatibility documentation for the exact ESXi version and hardware configuration. A server that works with one release may require a different driver, firmware level, or installation image for another release. An installer reporting unsupported hardware or no compatible devices commonly indicates an unsupported component or missing compatibility requirement.

Do not confuse physical-host requirements with guest operating system support. The host must meet ESXi's hardware requirements, while each VM must use a guest OS supported by the selected ESXi release and configured with appropriate virtual hardware. An ESXi host can run supported 32-bit and 64-bit guest operating systems when the ESXi version and compatibility guidance allow them.

Intel Xeon-class and AMD server processors are examples of processor families seen in server virtualization. They are not, by themselves, current compatibility requirements; verify the exact model and release documentation.

Capacity and Scalability

ESXi and vSphere releases publish configuration maximums for values such as logical CPUs, vCPUs, host memory, and the number of VMs per host. These maximums vary by release, licensing, hardware, configuration, and workload design.

A configuration maximum is a product or platform boundary, not a recommended workload target. A host can reach poor performance because of CPU contention, memory pressure, storage latency, or network congestion long before it reaches a published maximum. Consult the configuration maximums for the deployed ESXi and vSphere version, then perform workload-specific capacity planning and monitoring.

Installation Media and Deployment Footprint

ESXi can be installed on supported local disks and boot devices, as well as supported remote or embedded boot media depending on the release and server platform. Boot-device support, endurance requirements, and installation methods must follow the ESXi release guidance and the hardware vendor's recommendations.

ESXi has a comparatively small footprint and a limited-purpose design focused on virtualization. A focused design can reduce unnecessary components, simplify maintenance, and support reliability goals. Small size alone does not guarantee security: secure configuration, patching, access control, network design, and monitoring remain necessary.

For a deployment-oriented introduction, see ESXi Installation.

ESXi in the vSphere Ecosystem

ESXi is the host hypervisor layer. Other vSphere components can add centralized inventory, management, clustering, automation, and policy-based capabilities.

An individual ESXi host can be managed directly through supported host-management interfaces. vCenter Server is the centralized management platform used to manage multiple ESXi hosts and vSphere resources. Advanced capabilities may require vCenter Server, appropriate licensing, compatible storage or networking, and a suitable cluster design. ESXi alone does not automatically provide every vSphere feature.

See Add ESXi Host to vCenter Server Inventory and Create Clusters for related concepts.

Historical ESX and ESXi Distinction

VMware previously offered both ESX and ESXi architectures. The older ESX architecture included a service console, while ESXi used a more streamlined architecture with a smaller management footprint.

ESXi became the standard hypervisor architecture beginning with the vSphere 5 era. This distinction is historical context rather than deployment guidance: current planning should use the documentation and compatibility requirements for the ESXi release being deployed.

Local Host Administration with the DCUI

The Direct Console User Interface (DCUI) is the local console available from an ESXi server's screen and keyboard or an equivalent remote physical-console function. It is useful for initial configuration and recovery-oriented tasks when network management is not yet available.

At a high level, an administrator can use the DCUI to:

  • Select the management network adapter.
  • Set VLAN information when required.
  • Choose DHCP or static IPv4 and IPv6 addressing.
  • Configure DNS and hostname values.
  • Test management-network connectivity.
  • View host status.
  • Restart management agents when appropriate.
  • Enable controlled troubleshooting access according to organizational policy.

DCUI access is different from browser-based host management and from centralized vCenter Server management. After installation, an administrator might configure the management network in the DCUI, verify connectivity, and then add the host to centralized management if the environment requires it.

Common Troubleshooting Situations

Installer reports unsupported hardware or no compatible devices

  • Likely causes: the server, CPU, storage controller, or network adapter is unsupported for the selected ESXi release, or required firmware and driver compatibility is missing.
  • Approach: verify the exact hardware and firmware combination against the compatibility documentation before changing the design.

Installed host cannot be reached over the management network

  • Likely causes: incorrect IP address, subnet mask, gateway, DNS, VLAN, or physical uplink selection; alternatively, the switch port may not carry the expected management VLAN.
  • Approach: use the DCUI to inspect management-network values and test connectivity step by step, then check the physical switch configuration.

VMs perform poorly after more workloads are deployed

  • Likely causes: CPU, memory, storage, or network contention, or VM sizing that exceeds practical host capacity.
  • Approach: identify the constrained resource with monitoring and distinguish workload capacity from published product maximums.

Planned guest OS cannot be installed or run as expected

  • Likely causes: the guest OS version is unsupported by the selected ESXi release, or the VM configuration does not meet that guest OS's requirements.
  • Approach: check guest OS compatibility separately from physical-host compatibility and review the VM's virtual hardware configuration.

ESXi Terminology Quick Reference

TermMeaningWhy It Matters
HostPhysical server running ESXi.Supplies the finite hardware capacity shared by VMs.
Virtual machineSoftware-defined computer.Provides isolation and independent virtual hardware.
Guest operating systemOS installed inside a VM.Runs applications on virtual hardware.
VMkernelESXi kernel coordinating VM resource access.Provides scheduling, abstraction, and hardware mediation.
DatastoreStorage presented to ESXi for VM files.Holds virtual disks, configuration files, and related data.
vSwitchSoftware-based virtual switch.Connects VMs and physical network uplinks.
DCUILocal Direct Console User Interface.Supports initial and recovery-oriented host administration.
vCenter ServerCentralized management platform for ESXi hosts.Enables multi-host inventory and selected vSphere capabilities.

Key Takeaways

  • ESXi is VMware's Type 1 hypervisor and installs directly on compatible physical server hardware.
  • The VMkernel mediates VM access to CPU, memory, storage, and networking resources.
  • VMs share finite host capacity but maintain logical separation between guest operating systems and applications.
  • Type 2 products such as VMware Workstation and Oracle VirtualBox run inside a conventional host operating system and are commonly used on desktops and laptops.
  • Hardware compatibility and guest OS compatibility are separate checks.
  • Published configuration maximums are not the same as practical workload capacity.
  • ESXi can be managed directly, while vCenter Server centralizes management of multiple hosts and supports additional vSphere capabilities.