Merge pull request #10029 from MicrosoftDocs/main

Auto Publish – main to live - 2025-10-23 18:00 UTC

Author: learn-build-service-prod[bot]

support/azure/virtual-desktop/toc.yml

@@ -35,5 +35,7 @@ items:
   href: ./troubleshoot-session-host-configuration-update.md
 - name: Troubleshoot session host virtual machine configuration
   href: ./troubleshoot-vm-configuration.md
+- name: Troubleshoot ephemeral OS disks
+  href: ./troubleshoot-ephemeral-os-disks.md
 - name: Win key remains held after pressing Ctrl+Win+L in a remote session
   href: win-key-remains-held-after-pressing-ctrl-win-l-in-remote-session.md

support/azure/virtual-desktop/troubleshoot-ephemeral-os-disks.md

@@ -0,0 +1,88 @@
+---
+title: Troubleshoot ephemeral OS disks in Azure Virtual Machines (VMs) and Azure Virtual Desktop (AVD)
+description: Resolve deployment failures, data loss, and feature limitations with ephemeral OS disks for Azure Virtual Machines (VMs) and Azure Virtual Desktop (AVD). Get troubleshooting steps and best practices.
+ms.topic: troubleshooting
+ms.date: 10/23/2025
+ms.reviewer: kaushika
+audience: itpro
+ms.custom: 
+- sap:Azure Virtual Desktop\Remote Desktop Clients\Redirecting resources via the client
+- pcy:wincomm-user-experience
+
+---
+# Troubleshoot ephemeral OS disks in Azure Virtual Machines (VMs) and Azure Virtual Desktop (AVD)
+
+## Summary
+
+Ephemeral OS disks (EOSD) are designed for Azure Virtual Machines (VMs) and Azure Virtual Desktop (AVD) session hosts, offering faster provisioning and reduced storage costs. However, users might encounter deployment failures, unexpected data loss, or functional limitations when the stateless design of EOSD and its dependencies on VM size and local storage aren't fully understood. This article helps identify and resolve these issues.
+
+## Symptoms
+
+- Deployment failures occur because of insufficient local storage for the OS image on specific VM sizes.
+- Loss of OS-level changes after stopping, deallocating, or reimaging a VM.
+- Confusion arises from unsupported EOSD features like snapshots, backups, and encryption.
+- Operational delays and increased support cases for AVD and stateless workloads.
+
+## Troubleshooting Guidance
+
+### VM size incompatibility
+
+**Cause**: Some VM sizes don't have enough local cache or temporary disk space to accommodate the OS image, causing deployment failures.
+
+**Solution**:
+
+- Check that the VM size supports EOSD by verifying the local cache or temporary disk size.
+- Use the Azure CLI command `EphemeralOSDiskSupported` to confirm compatibility.
+- For AVD, ensure the Temp Disk placement is selected during VM creation, as NVMe is not supported in public preview.
+
+### Misunderstanding EOSD persistence
+
+**Cause**: Ephemeral OS disks are designed as stateless disks, which means they don't support persistent features like snapshots, backups, or encryption.
+
+**Solution**: Use managed OS disks for persistent features. If you need features such as backups, snapshots, or encryption, switch to managed OS disks instead of EOSD for your Azure VMs or AVD session hosts.
+
+### Lifecycle mismanagement
+
+**Cause**: Stop and deallocate operations wipe the OS state on EOSD, causing data loss.
+
+**Solution**: Avoid stop and deallocate operations.
+
+- Don't stop or deallocate VMs using EOSD, as this leads to data loss. Instead, design stateless workloads.
+- Persist user profiles externally using FSLogix profile containers or Azure Files.
+
+### Incorrect DiffDiskPlacement configuration
+
+**Cause**: Selecting an unsupported placement option, such as NVMe, can cause deployment issues.
+
+**Solution**: Correct DiffDiskPlacement configuration
+
+- For AVD session hosts, ensure the Temp Disk placement is selected during deployment.
+- Avoid selecting NVMe placement during public preview to prevent deployment issues.
+
+### Regional SKU rollout gaps
+
+**Cause**: Incomplete rollout of supported SKUs or backend bugs cause regional deployment issues.
+
+**Solution**: Address regional SKU rollout gaps
+
+- Monitor Azure announcements for updates on SKU availability.
+- Deploy to another region or select a different SKU if rollout gaps affect your deployment.
+
+### Best practices
+- **Validate VM size and placement:** Before deployment, ensure your VM’s local cache or temporary disk can accommodate the OS image size.
+- **Design stateless workloads:** Persist user profiles and application data externally using FSLogix profile containers or Azure Files.
+- **Avoid stop/deallocate operations:** Use restart or delete/recreate logic for autoscale scenarios to prevent data loss.
+- **Communicate limitations upfront:** Inform teams that EOSD doesn't support snapshots, backups, or Azure Disk Encryption.
+- **Monitor regional availability:** Check supported SKUs and regions using Azure documentation before deployment.
+- **Set up automated configuration:** Use VM extensions or scripts to reapply settings and applications after reimaging.
+- **For AVD environments:**
+    1. Use Temp Disk for ephemeral OS disk placement (NVMe isn't supported in public preview).
+    1. Persist user profiles with FSLogix profile containers.
+    1. Store application packages on Azure Files or Azure NetApp Files.
+    1. Set up autoscale rules to delete and recreate hosts instead of deallocating them.
+
+### More information
+
+For further details, refer to the following resources:
+- [Ephemeral OS disks for AVD session hosts](/azure/virtual-desktop/deploy/session-hosts/ephemeral-os-disks)
+- [Ephemeral OS disks - Frequently Asked Questions](/azure/virtual-machines/ephemeral-os-disks-faq)

support/azure/virtual-machines/linux/linux-hyperv-issue.md

@@ -67,15 +67,15 @@ Or
     1. Identify the file that disables the hv_netvsc driver and the corresponding line numbers by running the following command:
 
         ```bash
-        grep -nr "hv_netvsc" /etc/modprobe.d/
+        sudo grep -nr "hv_netvsc" /etc/modprobe.d/
         ```
 
     2. Modify the corresponding file and comment out or delete the hv_netvsc entries:
 
           :::image type="content" source="media/linux-hyperv-issue/hv-netvsc-disabled.png" alt-text="Screenshot that shows the possible configuration file contents used to disable network drivers.":::
 
         ```bash
-        vi /etc/modprobe.d/disable.conf
+        sudo vi /etc/modprobe.d/disable.conf
         ```
 
         > [!NOTE]
@@ -88,13 +88,13 @@ Or
     - For RHEL/SLES-based images
 
         ```bash
-        # dracut -f -v
+        sudo dracut -f -v
         ```
 
     - For Ubuntu/Debian-based images
 
         ```bash
-        # mkinitramfs -k -o /boot/initrd.img-$(uname -r)
+        sudo mkinitramfs -k -o /boot/initrd.img-$(uname -r)
         ```
 
 6. Reboot the VM.
@@ -104,19 +104,19 @@ Always take a backup of the original initial RAMdisk image to facilitate the rol
 - For RHEL-based images:
 
     ```bash
-    # cp /boot/initramfs-<kernelVersion>.img /boot/initramfs-<kernelVersion>.img.bak
+    sudo cp /boot/initramfs-<kernelVersion>.img /boot/initramfs-<kernelVersion>.img.bak
     ```
 
 - For SLES-based images:
 
     ```bash
-    # cp /boot/initrd-<kernelVersion> /boot/initrd-<kernelVersion>.bak
+    sudo cp /boot/initrd-<kernelVersion> /boot/initrd-<kernelVersion>.bak
     ```
 
 - For Ubuntu/Debian-based images:
 
     ```bash
-    # cp /boot/initrd.img-<kernelVersion> /boot/initrd.img-<kernelVersion>.bak
+    sudo cp /boot/initrd.img-<kernelVersion> /boot/initrd.img-<kernelVersion>.bak
     ```
 
 ### <a id="reenable-hv_netvsc-offline"></a>Solution 2: Enable Hyper-V network driver offline
@@ -154,7 +154,7 @@ If the issue continues even though the Hyper-V network driver is enabled, use on
 7. Generally, a NIC MAC address would only change if a NIC is deleted or added by the administrator or a NIC is updated in the backend. If network configuration via cloud-init isn't desired, and the `apply_network_config` parameter needs to be set to false, delete the */var/lib/cloud/instance/obj.pkl* file and reboot the system.
 
     ```bash
-    # rm /var/lib/cloud/instance/obj.pkl
+    sudo rm /var/lib/cloud/instance/obj.pkl
     ```
 
 8. Once the changes are applied, restart the system.
@@ -213,7 +213,7 @@ If the VM is inaccessible due to other Hyper-V drivers being disabled, use an of
     1. Run the following command to identify the file that disables the hv_utils, hv_vmbus, hv_storvsc, or hv_netvsc driver and the corresponding line number.
 
         ```bash
-        egrep -nr "hv_utils|hv_vmbus|hv_storvsc|hv_netvsc" /etc/modprobe.d/
+        sudo egrep -nr "hv_utils|hv_vmbus|hv_storvsc|hv_netvsc" /etc/modprobe.d/
         ```
 
     2. Modify the corresponding file and comment out or delete the hv_utils, hv_vmbus, hv_storvsc, or hv_netvsc entries. The entries will most commonly be any of the following (or both):
@@ -223,7 +223,7 @@ If the VM is inaccessible due to other Hyper-V drivers being disabled, use an of
         :::image type="content" source="media/linux-hyperv-issue/hv-disabled-example-2.png" alt-text="Screenshot that shows the possible configuration file contents used to disable kernel modules/drivers.":::
 
         ```bash
-        vi /etc/modprobe.d/disable.conf
+        sudo vi /etc/modprobe.d/disable.conf
         ```
 
     > [!IMPORTANT]
@@ -236,13 +236,13 @@ If the VM is inaccessible due to other Hyper-V drivers being disabled, use an of
     - For RHEL/SLES-based images
 
         ```bash
-        # dracut -f -v
+        sudo dracut -f -v
         ```
 
     - For Ubuntu/Debian-based images
 
         ```bash
-        # mkinitramfs -k -o /boot/initrd.img-$(uname -r)
+        sudo mkinitramfs -k -o /boot/initrd.img-$(uname -r)
         ```
 
 5. Once the changes are applied, use the `az vm repair restore` command to perform an automatic OS disk swap with the original VM and reboot the system.

support/sql/tools/replay-markup-language-utility.md

@@ -1,9 +1,9 @@
 ---
 title: Replay Markup Language Utilities
 description: This article discusses a group of tools that are used by support professionals to troubleshoot SQL Server.
-ms.date: 09/09/2025
+ms.date: 10/21/2025
 ms.custom: sap:SQL Server Management, Query and Data Tools
-ms.reviewer: sureshka, jopilov, toddhayn, troymoen, v-jayaramanp
+ms.reviewer: sureshka, jopilov, toddhayn, troymoen, v-jayaramanp, v-shaywood
 ---
 # Replay Markup Language (RML) Utilities for SQL Server
 
@@ -23,6 +23,9 @@ The RML Utilities are a set of diagnostic tools for troubleshooting and aiding p
 
 For a complete description of every tool and sample usage, see the RML Help file that's included in RML Utilities for SQL Server.
 
+> [!NOTE]
+> RML Utilities have been deprecated as of October 2025. This product is no longer in development and no future enhancements or fixes should be expected.
+
 ## How are RML Utilities useful?
 
 You can use RML Utilities for SQL Server to do the following tasks:
@@ -50,7 +53,7 @@ You can examine the version history of RML Utilities in this table and download
 
 |Version number|Description|
 |---|---|
-|[09.04.0103](https://www.microsoft.com/download/details.aspx?id=106287)| Indicates the current web release that's available from the Microsoft Download Center. It supports all released versions of SQL Server (SQL Server 2022, SQL Server 2019, SQL Server 2017, SQL Server 2016, SQL Server 2014, SQL Server 2012, SQL Server 2008 R2, and SQL Server 2008).|
+|[09.04.0103](https://www.microsoft.com/download/details.aspx?id=106287)| The current and final web release that's available from the Microsoft Download Center. It supports all released versions of SQL Server (SQL Server 2022, SQL Server 2019, SQL Server 2017, SQL Server 2016, SQL Server 2014, SQL Server 2012, SQL Server 2008 R2, and SQL Server 2008).|
 |[09.04.0102](https://www.microsoft.com/download/details.aspx?id=104868)| Indicates the previous web release that's available from the Microsoft Download Center. It supports all released versions of SQL Server (SQL Server 2022, SQL Server 2019, SQL Server 2017, SQL Server 2016, SQL Server 2014, SQL Server 2012, SQL Server 2008 R2, and SQL Server 2008).|
 |[9.04.0100](https://www.microsoft.com/download/details.aspx?id=103126)| Indicates a previous web release that's available from the Microsoft Download Center. It supports all released versions of SQL Server (SQL Server 2019, SQL Server 2017, SQL Server 2016, SQL Server 2014, SQL Server 2012, SQL Server 2008 R2, SQL Server 2008, SQL Server 2005, and SQL Server 2000).|
 |[9.04.0098](https://www.microsoft.com/download/details.aspx?id=54090)| Indicates a previous web release package that is included with the Database Experimentation Assistant tool. It supports all released versions of SQL Server.|

support/windows-365/how-to-deprovision-business-grace-with-graph-explorer.md

@@ -0,0 +1,94 @@
+---
+title: How to Remove Windows 365 Business Cloud PCs from Grace Using Graph Explorer and Microsoft Graph API
+description: Learn how to remove Windows 365 Business Cloud PCs from grace period using Microsoft Graph Explorer and API. Step-by-step guide with prerequisites and troubleshooting tips.
+manager: dcscontentpm
+ms.reviewer: kaushika, anwill
+ms.date: 10/22/2025
+ms.topic: troubleshooting
+ms.custom:
+- pcy:Provisioning\Grace Period Issues
+- sap:WinComm User Experience
+---
+# How to remove Windows 365 Business Cloud PCs from grace period using Graph Explorer and Microsoft Graph API
+
+## Summary
+
+When you remove a license from a user, Windows 365 Business Cloud PCs enter a grace period before permanent removal. In some cases, you might want to remove these Cloud PCs from grace manually, such as to reassign the license to another user.
+
+This article provides step-by-step instructions for using Microsoft Graph Explorer and the Microsoft Graph API to perform this deprovision.
+
+## Prerequisites
+
+- Administrator access to your Microsoft 365 tenant
+
+- Permissions to use Microsoft Graph API (Cloud PC administrator or equivalent)
+
+- Access to Microsoft Graph Explorer
+
+- Cloud PCs in `inGracePeriod` status in your Windows 365 environment
+
+## Step 1: Sign in to Graph Explorer
+
+1.  Go to [Graph Explorer \| Try Microsoft Graph APIs - Microsoft Graph](https://developer.microsoft.com/en-us/graph/graph-explorer)
+
+1.  Select **Sign in to Graph Explorer** (top right corner).
+
+1.  Authenticate by using your administrator credentials and set the tenant.
+
+## Step 2: Identify Cloud PCs in grace period
+
+1.  In the query box, set the following **GET** request:
+
+> https://graph.microsoft.com/beta/deviceManagement/virtualEndpoint/cloudPCs?\$filter=status eq 'inGracePeriod'
+
+1.  In Graph Explorer, select **Modify Permissions**.
+
+1.  Consent to the following permissions for your session:
+
+- CloudPC.Read.All
+
+- CloudPC.ReadWrite.All
+
+1.  If not already granted, select **Consent** to add these permissions.
+
+1.  Select **Run Query**.
+
+1.  Review the results to locate the Cloud PCs you want to remove. Note the **id** value for each relevant Cloud PC.
+
+## Step 3: Remove a Cloud PC from grace
+
+1.  For each Cloud PC, send a **POST** request to remove it from grace. In the query box, select **POST** and use the following request:
+
+> https://graph.microsoft.com/beta/deviceManagement/virtualEndpoint/cloudPCs/{cloudPCId}/endGracePeriod
+
+1.  Replace `{cloudPCId}` with the actual ID of the Cloud PC. For example:
+
+> https://graph.microsoft.com/beta/deviceManagement/virtualEndpoint/cloudPCs/4b18de4b-ab05-4059-8c61-0323a7df4ced/endGracePeriod
+
+1.  Leave the request body empty.
+
+1.  Select **Run Query**.
+
+1.  If successful, you receive a **No Content - 204** response.
+
+## Step 4: Verify removal
+
+1.  To confirm removal from grace, repeat the **GET** request from Step 3.
+
+1.  The Cloud PC shouldn't appear with status eq 'inGracePeriod'.
+
+## Troubleshooting and notes
+
+- If you receive a permissions error, verify that you have the correct roles and all necessary permissions are consented.
+
+- Changes made through the Graph API might take a few minutes to reflect in the Windows 365 portal.
+
+- Removing a Cloud PC from grace permanently deletes the Cloud PC and any associated user data.
+
+## Additional resources
+
+- Microsoft Graph Cloud PC Documentation
+
+  - [List cloudPCs - Microsoft Graph beta](/graph/api/virtualendpoint-list-cloudpcs?view=graph-rest-beta&tabs=http)
+
+  - [cloudPC: endGracePeriod - Microsoft Graph beta](/graph/api/cloudpc-endgraceperiod?view=graph-rest-beta&tabs=http)

support/windows-365/toc.yml

@@ -11,10 +11,10 @@ items:
     items:
     - name: Known issues with Windows 365 Business
       href: known-issues.md
-  - name: Resize
+  - name: Provisioning
     items:
-    - name: Resize a Windows 365 Business Cloud PC
-      href: resize-windows-365-business-cloud-pc.md
+    - name: Remove Windows 365 Business Cloud PCs from Grace Using Graph
+      href: how-to-deprovision-business-grace-with-graph-explorer.md
 - name: Windows 365 Enterprise
   items:
   - name: Azure Network Connection