Deploy a virtual Azure Stack HCI, version 23H2 system
Applies to: Azure Stack HCI, version 23H2
This article describes how to deploy a virtualized single server or a multi-node Azure Stack HCI, version 23H2, on a host system running Hyper-V on the Windows Server 2022, Windows 11, or later operating system (OS).
You need administrator privileges for the Azure Stack HCI virtual deployment and should be familiar with the existing Azure Stack HCI solution. The deployment can take around 2.5 hours to complete.
Important
A virtual deployment of Azure Stack HCI, version 23H2 is intended for educational and demonstration purposes only. Microsoft Support doesn't support virtual deployments.
Prerequisites
Here are the hardware, networking, and other prerequisites for the virtual deployment:
Physical host requirements
The following are the minimum requirements to successfully deploy Azure Stack HCI, version 23H2.
Before you begin, make sure that:
You have access to a physical host system that is running Hyper-V on Windows Server 2022, Windows 11, or later. This host is used to provision a virtual Azure Stack HCI deployment.
You have enough capacity. More capacity is required for running actual workloads like virtual machines or containers.
The physical hardware used for the virtual deployment meets the following requirements:
Component Minimum Processor Intel VT-x or AMD-V, with support for nested virtualization. For more information, see Does My Processor Support Intel® virtualization technology?. Memory The physical host must have a minimum of 32 GB RAM for single virtual node deployments. The virtual host VM should have at least 24 GB RAM.
The physical host must have a minimum of 64 GB RAM for two virtual node deployments. Each virtual host VM should have at least 24 GB RAM.Host network adapters A single network adapter. Storage 1 TB Solid state drive (SSD).
Virtual host requirements
Before you begin, make sure that each virtual host system can dedicate the following resources to provision your virtualized Azure Stack HCI system:
Component | Requirement |
---|---|
Virtual machine (VM) type | Secure Boot and Trusted Platform Module (TPM) enabled. |
vCPUs | Four cores. |
Memory | A minimum of 24 GB. |
Networking | At least two network adapters connected to internal network. MAC spoofing must be enabled. |
Boot disk | One disk to install the Azure Stack HCI operating system from ISO. |
Hard disks for Storage Spaces Direct | Six dynamic expanding disks. Maximum disk size is 1024 GB. |
Data disk | At least 127 GB. |
Time synchronization in integration | Disabled. |
Note
These are the minimum requirements to successfully deploy Azure Stack HCI, version 23H2. Increase the capacity like virtual cores and memory when running actual workloads like virtual machines or containers.
Set up the virtual switch
When deploying Azure Stack HCI in a virtual environment, you can use your existing networks and use IP addresses from that network if they're available. In such a case, you just need to create an external switch and connect all the virtual network adapters to that virtual switch. Virtual hosts will have connectivity to your physical network without any extra configuration.
However, if your physical network where you're planning to deploy the Azure Stack HCI virtual environment is scarce on IPs, you can create an internal virtual switch with NAT enabled, to isolate the virtual hosts from your physical network while keeping outbound connectivity to the internet.
The following lists the steps for the two options:
Deploy with external virtual switch
On your physical host computer, run the following PowerShell command to create an external virtual switch:
New-VMSwitch -Name "external_switch_name" -SwitchType External -NetAdapterName "network_adapter_name" -AllowManagementOS $true
Deploy with internal virtual switch and NAT enabled
On your physical host computer, run the following PowerShell command to create an internal virtual switch. The use of this switch ensures that the Azure Stack HCI deployment is isolated.
New-VMSwitch -Name "internal_switch_name" -SwitchType Internal -NetAdapterName "network_adapter_name"
Once the internal virtual switch is created, a new network adapter is created on the host. You must assign an IP address to this network adapter to become the default gateway of your virtual hosts once connected to this internal switch network. You also need to define the NAT network subnet where the virtual hosts are connected.
The following example script creates a NAT network HCINAT
with prefix 192.168.44.0/24
and defines the 192.168.44.1
IP as the default gateway for the network using the interface on the host:
#Check interface index of the new network adapter on the host connected to InternalSwitch:
Get-NetAdapter -Name "vEthernet (InternalSwitch)"
#Create the NAT default gateway IP on top of the InternalSwitch network adapter:
New-NetIPAddress -IPAddress 192.168.44.1 -PrefixLength 24 -InterfaceAlias "vEthernet (InternalSwitch)"
#Create the NAT network:
New-NetNat -Name "HCINAT"-InternalIPInterfaceAddressPrefix 192.168.44.0/24
Create the virtual host
Create a VM to serve as the virtual host with the following configuration. You can create this VM using either Hyper-V Manager or PowerShell:
Hyper-V Manager. For more information, see Create a virtual machine using Hyper-V Manager to mirror your physical management network.
PowerShell cmdlets. Make sure to adjust the VM configuration parameters referenced in the Virtual host requirements before you run the PowerShell cmdlets.
Follow these steps to create an example VM named Node1
using PowerShell cmdlets:
Create the VM:
New-VHD -Path "your_VHDX_path" -SizeBytes 127GB New-VM -Name Node1 -MemoryStartupBytes 20GB -VHDPath "your_VHDX_path" -Generation 2 -Path "VM_config_files_path"
Disable dynamic memory:
Set-VMMemory -VMName "Node1" -DynamicMemoryEnabled $false
Disable VM checkpoints:
Set-VM -VMName "Node1" -CheckpointType Disabled
Remove the default network adapter created during VM creation in the previous step:
Get-VMNetworkAdapter -VMName "Node1" | Remove-VMNetworkAdapter
Add new network adapters to the VM using custom names. This example adds four NICs, but you can add just two if needed. Having four NICs allows you to test two network intents (
Mgmt_Compute
andStorage
for example) with two NICs each:Add-VmNetworkAdapter -VmName "Node1" -Name "NIC1" Add-VmNetworkAdapter -VmName "Node1" -Name "NIC2" Add-VmNetworkAdapter -VmName "Node1" -Name "NIC3" Add-VmNetworkAdapter -VmName "Node1" -Name "NIC4"
Attach all network adapters to the virtual switch. Specify the name of the virtual switch you created, whether it was external without NAT, or internal with NAT:
Get-VmNetworkAdapter -VmName "Node1" |Connect-VmNetworkAdapter -SwitchName "virtual_switch_name"
Enable MAC spoofing on all network adapters on VM
Node1
. MAC address spoofing is a technique that allows a network adapter to masquerade as another by changing its Media Access Control (MAC) address. This is required in scenarios where you're planning to use nested virtualization:Get-VmNetworkAdapter -VmName "Node1" |Set-VmNetworkAdapter -MacAddressSpoofing On
Enable trunk port (for multi-node deployments only) for all network adapters on VM
Node1
. This script configures the network adapter of a specific VM to operate in trunk mode. This is typically used in multi-node deployments where you want to allow multiple Virtual Local Area Networks (VLANs) to communicate through a single network adapter:Get-VmNetworkAdapter -VmName "Node1" |Set-VMNetworkAdapterVlan -Trunk -NativeVlanId 0 -AllowedVlanIdList 0-1000
Create a new key protector and assign it to
Node1
. This is typically done in the context of setting up a guarded fabric in Hyper-V, a security feature that protects VMs from unauthorized access or tampering.After the following script is executed,
Node1
will have a new key protector assigned to it. This key protector protects the VM's keys, helping to secure the VM against unauthorized access or tampering:$owner = Get-HgsGuardian UntrustedGuardian $kp = New-HgsKeyProtector -Owner $owner -AllowUntrustedRoot Set-VMKeyProtector -VMName "Node1" -KeyProtector $kp.RawData
Enable the vTPM for
Node1
. By enabling vTPM on a VM, you can use BitLocker and other features that require TPM on the VM. After this command is executed,Node1
will have a vTPM enabled, assuming the host machine's hardware and the VM's configuration support this feature.Enable-VmTpm -VMName "Node1"
Change virtual processors to
8
:Set-VmProcessor -VMName "Node1" -Count 8
Create extra drives to be used as the boot disk and hard disks for Storage Spaces Direct. After these commands are executed, six new VHDXs will be created in the
C:\vms\Node1
directory as shown in this example:new-VHD -Path "C:\vms\Node1\s2d1.vhdx" -SizeBytes 1024GB new-VHD -Path "C:\vms\Node1\s2d2.vhdx" -SizeBytes 1024GB new-VHD -Path "C:\vms\Node1\s2d3.vhdx" -SizeBytes 1024GB new-VHD -Path "C:\vms\Node1\s2d4.vhdx" -SizeBytes 1024GB new-VHD -Path "C:\vms\Node1\s2d5.vhdx" -SizeBytes 1024GB new-VHD -Path "C:\vms\Node1\s2d6.vhdx" -SizeBytes 1024GB
Attach drives to the newly created VHDXs for the VM. In these commands, six VHDs located in the
C:\vms\Node1
directory and nameds2d1.vhdx
throughs2d6.vhdx
are added toNode1
. EachAdd-VMHardDiskDrive
command adds one VHD to the VM, so the command is repeated six times with different-Path
parameter values.Afterwards, the
Node1
VM has six VHDs attached to it. These VHDXs are used to enable Storage Spaces Direct on the VM, which are required for Azure Stack HCI deployments:Add-VMHardDiskDrive -VMName "Node1" -Path "C:\vms\Node1\s2d1.vhdx" Add-VMHardDiskDrive -VMName "Node1" -Path "C:\vms\Node1\s2d2.vhdx" Add-VMHardDiskDrive -VMName "Node1" -Path "C:\vms\Node1\s2d3.vhdx" Add-VMHardDiskDrive -VMName "Node1" -Path "C:\vms\Node1\s2d4.vhdx" Add-VMHardDiskDrive -VMName "Node1" -Path "C:\vms\Node1\s2d5.vhdx" Add-VMHardDiskDrive -VMName "Node1" -Path "C:\vms\Node1\s2d6.vhdx"
Disable time synchronization:
Get-VMIntegrationService -VMName "Node1" |Where-Object {$_.name -like "T*"}|Disable-VMIntegrationService
Enable nested virtualization:
Set-VMProcessor -VMName "Node1" -ExposeVirtualizationExtensions $true
Start the VM:
Start-VM "Node1"
Install the OS on the virtual host VMs
Complete the following steps to install and configure the Azure Stack HCI OS on the virtual host VMs:
Download Azure Stack HCI 23H2 ISO and Install the Azure Stack HCI operating system.
Update the password since this is the first VM startup. Make sure the password meets the Azure complexity requirements. The password is at least 12 characters and includes 1 uppercase character, 1 lowercase character, 1 number, and 1 special character.
After the password is changed, the Server Configuration Tool (SConfig) is automatically loaded. Select option
15
to exit to the command line and run the next steps from there.Launch SConfig by running the following command:
SConfig
For information on how to use SConfig, see Configure with the Server Configuration tool (SConfig).
Change hostname to
Node1
. Use option2
forComputer name
in SConfig to do this.The hostname change results in a restart. When prompted for a restart, enter
Yes
and wait for the restart to complete. SConfig is launched again automatically.From the physical host, run the
Get-VMNetworkAdapter
andForEach-Object
cmdlets to configure the four network adapter names for VMNode1
by mapping the assigned MAC addresses to the corresponding network adapters on the guest OS.- The
Get-VMNetworkAdapter
cmdlet is used to retrieve the network adapter object for each NIC on the VM, where the-VMName
parameter specifies the name of the VM, and the-Name
parameter specifies the name of the network adapter. TheMacAddress
property of the network adapter object is then accessed to get the MAC address:
Get-VMNetworkAdapter -VMName "Node1" -Name "NIC1"
The MAC address is a string of hexadecimal numbers. The
ForEach-Object
cmdlet is used to format this string by inserting hyphens at specific intervals. Specifically, theInsert
method of the string object is used to insert a hyphen at the 2nd, 5th, 8th, 11th, and 14th positions in the string. Thejoin
operator is then used to concatenate the resulting array of strings into a single string with spaces between each element.The commands are repeated for each of the four NICs on the VM, and the final formatted MAC address for each NIC is stored in a separate variable:
($Node1finalmacNIC1, $Node1finalmacNIC2, $Node1finalmacNIC3, $Node1finalmacNIC4).
- The following script outputs the final formatted MAC address for each NIC:
$Node1macNIC1 = Get-VMNetworkAdapter -VMName "Node1" -Name "NIC1" $Node1macNIC1.MacAddress $Node1finalmacNIC1=$Node1macNIC1.MacAddress|ForEach-Object{($_.Insert(2,"-").Insert(5,"-").Insert(8,"-").Insert(11,"-").Insert(14,"-"))-join " "} $Node1finalmacNIC1 $Node1macNIC2 = Get-VMNetworkAdapter -VMName "Node1" -Name "NIC2" $Node1macNIC2.MacAddress $Node1finalmacNIC2=$Node1macNIC2.MacAddress|ForEach-Object{($_.Insert(2,"-").Insert(5,"-").Insert(8,"-").Insert(11,"-").Insert(14,"-"))-join " "} $Node1finalmacNIC2 $Node1macNIC3 = Get-VMNetworkAdapter -VMName "Node1" -Name "NIC3" $Node1macNIC3.MacAddress $Node1finalmacNIC3=$Node1macNIC3.MacAddress|ForEach-Object{($_.Insert(2,"-").Insert(5,"-").Insert(8,"-").Insert(11,"-").Insert(14,"-"))-join " "} $Node1finalmacNIC3 $Node1macNIC4 = Get-VMNetworkAdapter -VMName "Node1" -Name "NIC4" $Node1macNIC4.MacAddress $Node1finalmacNIC4=$Node1macNIC4.MacAddress|ForEach-Object{($_.Insert(2,"-").Insert(5,"-").Insert(8,"-").Insert(11,"-").Insert(14,"-"))-join " "} $Node1finalmacNIC4
- The
Obtain the
Node1
VM local admin credentials and then renameNode1
:$cred = get-credential
Rename and map the NICs on
Node1
. The renaming is based on the MAC addresses of the NICs assigned by Hyper-V when the VM is started the first time. These commands should be run directly from the host:Use the
Get-NetAdapter
command to retrieve the physical network adapters on the VM, filter them based on their MAC address, and then rename them to the matching adapter using theRename-NetAdapter
cmdlet.This is repeated for each of the four NICs on the VM, with the MAC address and new name of each NIC specified separately. This establishes a mapping between the name of the NICs in Hyper-V Manager and the name of the NICs in the VM OS:
Invoke-Command -VMName "Node1" -Credential $cred -ScriptBlock {param($Node1finalmacNIC1) Get-NetAdapter -Physical | Where-Object {$_.MacAddress -eq $Node1finalmacNIC1} | Rename-NetAdapter -NewName "NIC1"} -ArgumentList $Node1finalmacNIC1 Invoke-Command -VMName "Node1" -Credential $cred -ScriptBlock {param($Node1finalmacNIC2) Get-NetAdapter -Physical | Where-Object {$_.MacAddress -eq $Node1finalmacNIC2} | Rename-NetAdapter -NewName "NIC2"} -ArgumentList $Node1finalmacNIC2 Invoke-Command -VMName "Node1" -Credential $cred -ScriptBlock {param($Node1finalmacNIC3) Get-NetAdapter -Physical | Where-Object {$_.MacAddress -eq $Node1finalmacNIC3} | Rename-NetAdapter -NewName "NIC3"} -ArgumentList $Node1finalmacNIC3 Invoke-Command -VMName "Node1" -Credential $cred -ScriptBlock {param($Node1finalmacNIC4) Get-NetAdapter -Physical | Where-Object {$_.MacAddress -eq $Node1finalmacNIC4} | Rename-NetAdapter -NewName "NIC4"} -ArgumentList $Node1finalmacNIC4
Disable the Dynamic Host Configuration Protocol (DHCP) on the four NICs for VM
Node1
by running the following commands.Note
The interfaces won't automatically obtain IP addresses from a DHCP server and instead need to have IP addresses manually assigned to them:
Invoke-Command -VMName "Node1" -Credential $cred -ScriptBlock {Set-NetIPInterface -InterfaceAlias "NIC1" -Dhcp Disabled} Invoke-Command -VMName "Node1" -Credential $cred -ScriptBlock {Set-NetIPInterface -InterfaceAlias "NIC2" -Dhcp Disabled} Invoke-Command -VMName "Node1" -Credential $cred -ScriptBlock {Set-NetIPInterface -InterfaceAlias "NIC3" -Dhcp Disabled} Invoke-Command -VMName "Node1" -Credential $cred -ScriptBlock {Set-NetIPInterface -InterfaceAlias "NIC4" -Dhcp Disabled}
Set management IP, gateway, and DNS. After the following commands are executed,
Node1
will have theNIC1
network interface configured with the specified IP address, subnet mask, default gateway, and DNS server address. Ensure that the management IP address can resolve Active Directory and has outbound connectivity to the internet:Invoke-Command -VMName "Node1" -Credential $cred -ScriptBlock {New-NetIPAddress -InterfaceAlias "NIC1" -IPAddress "192.168.44.201" -PrefixLength 24 -AddressFamily IPv4 -DefaultGateway "192.168.44.1"} Invoke-Command -VMName "Node1" -Credential $cred -ScriptBlock {Set-DnsClientServerAddress -InterfaceAlias "NIC1" -ServerAddresses "192.168.1.254"}
Enable the Hyper-V role. This command restarts the VM
Node1
:Invoke-Command -VMName "Node1" -Credential $cred -ScriptBlock {Enable-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V -All }
Once
Node1
is restarted and the Hyper-V role is installed, install the Hyper-V Management Tools:Invoke-Command -VMName "Node1" -Credential $cred -ScriptBlock {Install-WindowsFeature -Name Hyper-V -IncludeManagementTools}
Once the virtual host server is ready, you must register it and assign permissions in Azure as an Arc resource.
Once the server is registered in Azure as an Arc resource and all the mandatory extensions are installed, choose one of the following methods to deploy Azure Stack HCI from Azure.
Repeat the process above for extra nodes if you plan to test multi-node deployments. Ensure virtual host names and management IPs are unique and on the same subnet: