Unable to delete Azure Databricks workspace (stuck in 'Deleting' state)

Daria Kolosovska 0 Reputation points
2026-03-17T16:01:58.28+00:00

I am experiencing an issue with an Azure Databricks workspace that is stuck in the "Deleting" state, and I would appreciate your assistance.

I created a Databricks workspace along with its managed resource group for a learning task. After completing my work, I attempted to delete the workspace. However, it has now been in the "Deleting" state for over 20 hours without any progress.

At this point:

I am unable to cancel the deletion process and retry

The "Open workspace" button is not available, so I cannot access the workspace to verify whether any resources or jobs are still running

There are no resource locks on the workspace or its resource group

There are no Deny Assignments (except the expected system-managed ones on the managed resource group)

I attempted deletion via Azure CLI using the "-force" flag, but received the error: "Code: ApplianceBeingDeleted"

Additionally, I cannot delete the managed resource group due to: "Code: DenyAssignmentAuthorizationFailed", as it can only be deleted by the Databricks resource itself.

Currently, I am blocked from cleaning up these resources and I am concerned about potential ongoing charges for infrastructure that I cannot control or delete.

Could you please advise on how to resolve this issue or assist with forcefully removing the workspace and its associated resources?

Thank you in advance for your help.

Azure Databricks
Azure Databricks

An Apache Spark-based analytics platform optimized for Azure.

0 comments

1 answer

Sort by: Oldest
Most helpful Newest Oldest

  1. AI answer

    2026-03-17T16:02:24.9666667+00:00

    Azure Databricks workspace deletion is handled by the Azure control plane and Databricks; when a workspace is deleted it normally goes into a soft-deleted state for 7 days, but the Azure resource itself should transition to a completed “Deleted” state rather than being stuck.

    From the available information, there is no supported way to cancel a deletion or “force” it from the customer side once the resource is in a stuck Deleting state. The documented deletion and force-deletion options (Azure portal, Azure CLI, Azure PowerShell, and resource-group deletion) all assume that the control plane is responsive and able to complete the operation.

    The supported options are:

    1. Use the documented delete paths (which have already effectively been tried):
      • Azure portal: delete the workspace from the Databricks resource blade.
      • Azure CLI: az databricks workspace delete with --force-deletion if the workspace catalog and managed resource group need to be removed.
      • Azure PowerShell: Remove-AzDatabricksWorkspace -ForceDeletion.
      • Resource group deletion with force deletion types:
        • Azure CLI: az group delete --name <NameOfResourceGroup> --force-deletion-types Microsoft.Databricks/workspaces
        • Azure PowerShell: Remove-AzResourceGroup -Name <NameOfResourceGroup> -ForceDeletionType Microsoft.Databricks/workspaces -Force
      In the described state (“ApplianceBeingDeleted” and deny assignment blocking managed resource group deletion), these are already failing, which indicates a backend issue rather than a missing client-side step.
    2. Rely on the platform’s retention and cleanup behavior:
      • When a workspace is deleted, Azure Databricks-managed workspace resources (notebooks, jobs, workspace configuration) are subject to a 7‑day soft-delete retention window, but Azure cloud resources such as VMs, managed disks, and networking are cleaned up immediately when the workspace is deleted and are not subject to that retention.
      • This means that even while the workspace resource appears stuck in Deleting, the underlying compute resources that would accrue most of the cost should already be cleaned up by the platform.
    3. Engage Azure support to clear the stuck deletion:
      • Because the resource is in a control-plane state (ApplianceBeingDeleted) that prevents both further updates and deletion of the managed resource group (due to deny assignments that only the Databricks resource can remove), resolution requires backend intervention by the Azure/Databricks service teams.
      • The documented tools do not provide a customer-exposed way to override or remove those deny assignments or to complete the deletion when the service is stuck.

    In summary, all customer-exposed mechanisms for deleting or force-deleting an Azure Databricks workspace and its managed resource group have been exhausted in this scenario. The next step is to open a support case with Azure so the service team can fix the stuck deletion state and remove the workspace and its managed resource group from the subscription. Under normal operation, compute resources are already cleaned up at delete time, so ongoing infrastructure charges should be limited while the control-plane resource remains in Deleting.

    References:

    AI-generated content may be incorrect. Read our transparency notes for more information.

    Was this answer helpful?

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.