Upgrade Azure Stack HCI OS Version 22H2 or 3H2 to 24H2-OS Only

Document Overview

This article provides:

• An overview of an Azure Stack HCI OS only upgrade from version 22H2 or 23H2 to version 24H2
• Guidance on how to upgrade the Azure Stack HCI operating system (OS) to version 26100.xxxx (24H2) using PowerShell.
• There are two upgrade paths available:
  • Upgrade from version 20349.xxxx (22H2) to version 26100.xxxx (24H2).
  • Upgrade from version 25398.xxxx (23H2) to version 26100.xxxx (24H2). This includes the 25398.xxxx (23H2) clusters that were upgraded from version 20349.xxxx (22H2) and the solution upgrade is not yet applied.

With the 2505 release, a direct upgrade path from version 20349.xxxx (22H2) to version 26100.xxxx (24H2) is available. Going directly from OS version 20349.xxxx (22H2) to 26100.xxxx (24H2) helps reduce reboots and maintenance planning prior to the solution upgrade.

Note: 22H2 is officially end of life as of May 31, 2025.

Full (OS+Solution) upgrades from 22H2 to 23H2, 22H2 to 24H2, or 23H2 (OS upgraded only) to 24H2 are a PAID service from Dell and are NOT covered by a support contract. Microsoft is offering limited assistance with upgrade-related issues for cases opened directly with them through the Azure Portal. https://learn.microsoft.com/en-us/azure/azure-local/upgrade/about-upgrades-23h2

Prerequisites

Before you begin the upgrade:

1. Ensure that your Azure Local system is running either OS version 20349.3692 or OS version greater than 25398.1611.
2. Ensure that all the machines in the cluster are healthy and show as online and registered to Azure.
3. Ensure that your region has Azure Stack HCI OS, version 23H2, available. Upgrading your Azure Local OS from version 22H2 is only supported in some regions. For more information, see Azure Requirements  
4. Ensure to shut down virtual machines (VMs). Microsoft recommends shutting down VMs before performing the OS upgrade to prevent unexpected outages and damage to databases.
5. Ensure that the Cluster-Aware Updating (CAU) role is installed and enabled on the cluster, see Add-CauClusterRole
6. Ensure that the Network Interface Card (NIC) driver installed on your system is newer than the version included by default (inbox) with Azure Stack HCI OS, version 26100.xxxx. Use the below cmdlet to verify.

Get-WindowsDriver -Online | Where-Object {$_.ClassName -like "net*" -and $_.Inbox -eq $true}

7. Ensure a Trusted Platform Module 2.0 (TPM 2.0) is installed on all cluster machines and Secure Boot is enabled. 
8. Update the Registry Keys on each cluster machine. For more information, see Update Registry Keys

Steps to upgrade from version 22H2 or 23H2 to version 24H2.

To upgrade Azure Local OS from version 22H2 or 23H2 to 24H2, do the following multiphase steps:

1. Use the PowerShell Method to upgrade the OS
2. Do postoperating system upgrade steps
3. Install and enable Network ATC
4. Do solution upgrade:
   1. Validate solution upgrade readiness
   2. Apply solution upgrade

1. OS Upgrade using PowerShell

• PowerShell (Microsoft supported) - See Install new OS using PowerShell

2. Do postoperating system upgrade steps

For the detailed procedure, see Perform post operating system upgrade steps on Azure Local using PowerShell  

3. Install and enable Network ATC

If Network ATC is not configured, after you have upgraded the operating system from version 22H2 or 23H2 to version 24H2, you must set it up. For the detailed procedure, see: Configure Network ATC on Azure Local  

4. Do solution upgrade
   1. Validate solution upgrade readiness

After the Azure Local OS is upgraded from version 22H2 to version 24H2, Dell Technologies recommends that you validate the upgrade readiness of the cluster. For the detailed procedure, see Validate solution upgrade readiness for Azure Local  

2. Apply solution upgrade

After completion of the prechecks and cluster readiness, to proceed with the solution upgrade, see Install solution upgrade on Azure Local

Known issues and workarounds

• Network ATC intent shows PhysicalAdapterNotFound after the "OS upgrade" to 24H2

Workaround: See Network ATC intent health state

• A known issue results in VM Live Migrations failing during an OS upgrade

Workaround: Before doing an Azure Local OS upgrade, see Mitigation Details

• Test Cluster validation fails after the OS upgrade to 24H2, as remote NDIS shows as #2

$regPath = "HKLM:\system\currentcontrolset\services\clussvc\parameters"
$propertyName = "ExcludeAdaptersByDescription"
$propertyValue = "Remote NDIS Compatible Device,Remote NDIS Compatible Device #2"
New-ItemProperty -Path $regPath -Name $propertyName -Value $propertyValue -PropertyType String -Force

• Azure Stack registration status shows Out of Policy after the base OS upgrade from 22H2 or 23H2 to 24H2 on cluster nodes

1. 1. Run the following commands on one of the cluster nodes:

$Subscription = "Enter the Subscription ID"
$Tenant = "Enter the Tenant ID"
$RG = “name of the resource group”
$Region = "Region_name"
Register-AzStackHCI -SubscriptionId $Subscription -TenantId $Tenant -Region $Region – ResourceGroupName $RG -RepairRegistration

2. Do the following steps by starting with the first node, and then repeating on each subsequent node:
   1. To update $NodeName for each node, run the following command: 

$NodeName = "Nodexxx"

2. To drain roles from the node to ensure safe removal and verify that all roles have been moved off the node, run the following command:

Suspend-ClusterNode -Name $NodeName -Drain
Get-ClusterGroup

3. To evict the node from the cluster, run the following command: 

Remove-ClusterNode -Name $NodeName -Force

4. To reboot the node, run the following command: 

Restart-Computer -ComputerName $NodeName -Force

5. Check the operational status of all virtual disks and physical disks are in an OK state. Ensure that the storage repair job was completed by looking for enclave repair success, run the following commands: 

Get-WinEvent "Microsoft-AzureStack-Hci/Debug" -Oldest
Get-WinEvent "Microsoft-AzureStack-Hci/Debug" -Oldest | ? {$_.Message -match "Enclave"} 
Get-WinEvent "Microsoft-AzureStack-Hci/Debug" -Oldest | ? {$_.Message -match "Enclave repair complete."}

6. After repair, to readd the node to the cluster and resume the node back into service, run the following commands: 

Add-ClusterNode -Name $NodeName
Resume-ClusterNode -Name $NodeName

7. Repeat the procedure for the remaining nodes. 
8. Rerun the following commands on one of the cluster nodes and recheck the Azure Local (Azure Stack HCI) registration status.

$Subscription = "Enter the Subscription ID"
$Tenant = "Enter the Tenant ID"
$Region = "Region_name"
Register-AzStackHCI -RepairRegistration -SubscriptionId $Subscription -TenantId $Tenant -Region $Region
Get-AzureStackHCI | Format-List RegistrationStatus, ConnectionStatus, LastConnected, NextSync

For more information, see Out of policy error  
For a list of known issues and workarounds from Microsoft, see AzureLocal-Supportability/TSG/Upgrade at main · Azure/AzureLocal-Supportability · GitHub