Azure Databricks test workspaces stuck in "Deleting" state

Question

Hello Microsoft Community,

I have an issue with two Azure Databricks workspaces used for testing that have been stuck in the "Deleting" state for several days, and I cannot delete the associated Resource Groups.

Details:

• Workspace Names: PROD-DATABRICKS-WKS, DEV-DATABRICKS-WKS
• Resource Groups: PRO-DATABRICKS, DEV-DATABRICKS
• Location: East US
• Managed Resource Groups: automatically created by Databricks
• State: Deleting
• All clusters and jobs have been terminated
• There are no manual locks applied on the workspaces or resource groups (verified with Get-AzResourceLock in PowerShell)
• Previous deletion attempts via:
  • Azure Portal
  • PowerShell (Remove-AzDatabricksWorkspace)
  • Azure CLI (az databricks workspace delete)
  have all failed

Error message received:

"The operation cannot be performed on the appliance 'PROD-DATABRICKS-WKS/DEV-DATABRICKS-WKS' because it is being deleted (Code: ApplianceBeingDeleted)"

I am on a Basic support plan, so I cannot create a paid support ticket.

Questions:

1. Is there any free way to force the deletion of these Databricks test workspaces stuck in "Deleting"?
2. Can the community suggest a safe workaround to clear these workspaces and their Managed Resource Groups without a paid support plan?
3. Are there any known tricks or PowerShell/CLI commands that can help release these test workspaces from this stuck state?

These are non-production test environments used for practice, so I just want to clean them up safely.

Thank you in advance!

Answer 1

Hi Denis Guzman,
it looks like you’ve run into the classic “Deleting” limbo for Databricks workspaces and their managed RGs. Normally the async cleanup takes just a few minutes, but sometimes things hang if the backend catalog or RP operations get stuck. Here’s what you can try—100% free via Azure CLI or PowerShell, no paid support needed:

1. Verify the stuck state and dependencies • Run az databricks workspace show --resource-group <RG> --name <Workspace> to confirm the provisioningState is still Deleting. • List any resources in the managed RG that might still be cleaning up: az resource list --resource-group <ManagedRG> • Make sure there are no private endpoints or network interfaces still attached.
2. Force-delete the workspace + its managed RG (CLI) • Azure CLI has a built-in flag to force deletion of the workspace catalog and MRG:

az databricks workspace delete \
        --resource-group <YourRG> \
        --name <YourWorkspace> \
        --force-deletion \
        --yes
      ```  

• This will purge the workspace catalog (including Unity Catalog data if enabled) and then clean up the MRG.

3. Or force-delete via Resource Group deletion (CLI) If you just want to blow away the RG and the workspace together, use:

   az group delete \
     --name <YourRG> \
     --yes \
     --no-wait \
     --force-deletion-types Microsoft.Databricks/workspaces

   This tells ARM to forcibly tear down any lingering Databricks workspaces inside the RG.

4. PowerShell equivalent  


```powershell
 ```powershell
   # Force delete a workspace
   Remove-AzDatabricksWorkspace `
     -ResourceGroupName <YourRG> `
     -Name <YourWorkspace> `
     -ForceDeletion
   # Or force delete the RG + Databricks workspaces in it
   Remove-AzResourceGroup `
     -Name <YourRG> `
     -ForceDeletionType Microsoft.Databricks/workspaces `
     -Force

5. If the RG itself is stuck in “Deleting”  

   You can cancel a pending RG deletion and then retry the force delete:  


```bash
   az group cancel-deletion --name <YourRG>

   Then repeat step 2 or 3.

Reference docs  

• How to delete a workspace (includes force-delete flags)  

  https://learn.microsoft.com/azure/databricks/admin/workspace/delete-workspace#force-delete-the-workspace-catalog  

• Troubleshoot DeploymentBeingDeleted errors  

  https://learn.microsoft.com/cli/azure/databricks/workspace?view=azure-cli-latest#az-databricks-workspace-delete  

• Troubleshoot ResourceGroupBeingDeleted errors  

  https://learn.microsoft.com/azure/role-based-access-control/deny-assignments?tabs=azure-portal    
  
Hope this helps. If you have any follow-up questions, please let me know. I would be happy to help.

**Please do not forget to  "Accept Answer" and "up-vote" wherever the information provided helps you, as this can be beneficial to other community members.**

Answer 2

There is no supported way in Azure Databricks or Azure Resource Manager for customers to “force” a stuck workspace out of the Deleting state beyond the documented delete/force-delete options. When those fail and the resource remains in Deleting, the only resolution path is intervention by Microsoft support.

From the available information:

1. Force deletion options (when the workspace is still manageable)

If a workspace is not yet stuck and can still be managed, the supported “force delete” options are:

• Azure portal: when deleting the workspace, select the checkbox to delete the workspace along with its managed resource group. This force-deletes the workspace catalog and all associated data.
• PowerShell:

    Remove-AzDatabricksWorkspace -ForceDeletion -Name <NameOfWorkspace> -ResourceGroupName <NameOfResourceGroup>
  

• Azure CLI:

    az databricks workspace delete \
      --force-deletion \
      --name <NameOfWorkspace> \
      --resource-group <NameOfResourceGroup>
  

• Deleting the parent resource group with force deletion types:
  • Azure CLI:

        az group delete --name <NameOfResourceGroup> --force-deletion-types Microsoft.Databricks/workspaces
    

  • PowerShell:

        Remove-AzResourceGroup -Name <NameOfResourceGroup> -ForceDeletionType Microsoft.Databricks/workspaces -Force
    

These are the only documented “force” mechanisms. They permanently remove the workspace catalog, managed resource group, and all associated data.

2. Behavior when Unity Catalog is enabled

If Unity Catalog is auto-enabled on the workspace, deleting the workspace (without force deletion) retains the default workspace catalog and its storage:

• The managed resource group is converted to a regular resource group.
• The storage container holding Unity Catalog data and the access connector are retained.

To completely remove those, the force-deletion options above must be used.

3. When the workspace is already stuck in Deleting

Once the workspace is in a long-lived Deleting state and all documented delete/force-delete commands (portal, PowerShell, CLI, resource-group delete with --force-deletion-types) have been tried and still fail with an error such as ApplianceBeingDeleted, there is no additional public PowerShell/CLI “trick” to clear the state. At that point, the resource is in a backend-inconsistent state similar to other Azure resources that have been reported stuck in Creating/Deleting, which historically required internal engineering or support intervention to fix.

Because of that, there is no community-side, script-based, or free tooling-based workaround documented that can reliably clear a Databricks workspace stuck in Deleting. The resolution requires Microsoft to fix the backend state.

For a Basic support plan, the only viable next step is to open a support request through the Azure portal under the “technical” or “subscription management” categories that are available to Basic, or to work with the subscription owner/administrator to raise a support ticket so that the Databricks workspace and its resource groups can be cleaned up by Microsoft.

---

References:

• Delete a workspace
• az databricks workspace delete