Request to Enable Serverless Compute in Azure Databricks Workspace

Karthick E 0 Reputation points
2025-10-11T17:31:34.14+00:00

We have successfully provisioned the Azure Databricks workspace and its dependent resources (VNet, subnets, storage, and Key Vault) through Terraform in accordance with our HITRUST-compliant setup.

The next step is to enable Serverless Compute for the Databricks workspace, so that we can utilize Unity Catalog and SQL/Notebook workloads in a fully managed, autoscaling compute mode.

Could you please enable Serverless Compute in the Databricks workspace and ensure that our service principal has the necessary roles (Contributor + Databricks Admin) to manage this feature?

Once this is done, we can validate by running test notebooks and confirm the compute pool provisioning within Unity Catalog.

Azure Databricks
Azure Databricks
An Apache Spark-based analytics platform optimized for Azure.
{count} votes

2 answers

Sort by: Most helpful
  1. Kalyani Kondavaradala 3,145 Reputation points Microsoft External Staff Moderator
    2025-10-12T08:38:38.3666667+00:00

    Hi Karthick E,

    Thank you for reaching on Microsoft Q&A!

    you're looking to enable Serverless Compute in your Azure Databricks workspace.

    Here's how you can proceed:

    1. Ensure Eligibility

    Make sure that your workspace meets the following requirements to use Serverless Compute:

    • It must have Unity Catalog enabled.
    • It should be located in a supported region for serverless compute.
    • Your workspace must not have PCI-DSS enabled in its compliance profile.
    1. Check Workspace Plan

    You will need to be on the Premium plan to utilize Serverless SQL warehouses. If you're currently on a trial or standard plan, you’ll need to upgrade. You can do this through the Azure Portal.

    • Your Azure Databricks workspace must be on the Premium plan.
    1. Enable Serverless SQL

    To enable Serverless SQL for your workspace (if eligible):

    • You should automatically have serverless SQL warehouses enabled, but you can verify this under your workspace settings.
    • Your workspace must be in a region that supports Databricks SQL Serverless.
    • In case of issues, ensure that there are no restrictions in your compute policy that may prevent the usage of serverless compute.
    1. Role Assignments

    To ensure your service principal has the necessary access, verify that it has the following roles assigned:

    • Contributor
    • Databricks Admin
    1. Testing

    Once enabled, you can create and run test notebooks to confirm that the serverless compute pool is working correctly within Unity Catalog. If you encounter issues, the following steps might help:

    • Ensure that your Azure Storage firewall rules allow access from serverless compute nodes.
    • If you’re facing quota issues, you might need to request an increase in your serverless quota.

    Reference:

    https://free.blessedness.top/en-us/azure/databricks/compute/serverless

    https://free.blessedness.top/en-us/azure/databricks/admin/sql/serverless#limitations

    I hope this helps, let us know if you need further assistance. Kindly consider upvoting if the information is helpful.

    Thanks!

    Kalyani

    0 comments No comments

  2. PRADEEPCHEEKATLA 91,321 Reputation points Moderator
    2025-10-13T15:10:54.6+00:00

    Karthick E •Thanks for the question and using MS Q&A platform.

    You can enable or disable Serverless Compute for workflows and Notebooks directly for the select Previews option as shown below.

    User's image

    Note: This experience is only available to workspace and account admins. Account admins can manage Databricks account-level Previews from the Account console.

    For more details, refer to Manage Azure Databricks Previews and Enable serverless compute.


    𝘛𝘰 𝘴𝘵𝘢𝘺 𝘪𝘯𝘧𝘰𝘳𝘮𝘦𝘥 𝘢𝘣𝘰𝘶𝘵 𝘵𝘩𝘦 𝘭𝘢𝘵𝘦𝘴𝘵 𝘶𝘱𝘥𝘢𝘵𝘦𝘴 𝘢𝘯𝘥 𝘪𝘯𝘴𝘪𝘨𝘩𝘵𝘴 𝘰𝘯 𝘈𝘻𝘶𝘳𝘦 𝘋𝘢𝘵𝘢𝘣𝘳𝘪𝘤𝘬𝘴, 𝘥𝘢𝘵𝘢 𝘦𝘯𝘨𝘪𝘯𝘦𝘦𝘳𝘪𝘯𝘨, 𝘢𝘯𝘥 𝘤𝘭𝘰𝘶𝘥 𝘪𝘯𝘯𝘰𝘷𝘢𝘵𝘪𝘰𝘯𝘴, 𝘧𝘰𝘭𝘭𝘰𝘸 𝘮𝘦 𝘰𝘯 𝘓𝘪𝘯𝘬𝘦𝘥𝘐𝘯.

    0 comments No comments

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.