Azure Databricks - Stopped App Auto Started & Consumed Resources without any User Trigger

Damien Verrinder 20 Reputation points
2026-06-20T21:11:37.6166667+00:00

We have been automatically charged >$1,530 AUD in Azure resource costs for a Databricks app that was never switched on and has remained de-activated/dormant for at least 6 months prior (this was an old demo). This was last used in January 2025.

Without logging into the the Azure Management portal or the Databricks instance for this account for > 6 months, the service has automatically started the app in January 2026 and has consumed resources without any trigger or notification from our end. This has come to our attention through significant charges hitting our account.

We have tried raising as support ticket in Azure Management portal under this account, but we are not being presented with an option to get in contact with support.

We need a support case raised to investigate this error and reverse the charges.

Can Microsoft Support Help direct this?

Azure Databricks
Azure Databricks

An Apache Spark-based analytics platform optimized for Azure.

0 comments No comments

Answer accepted by question author
kagiyama yutaka 5,490 Reputation points
2026-06-21T12:15:48.9266667+00:00

I think the cleanest way is to open Help and support → Billing, pick the subscription, and send a charge‑review request with the Databricks charge dates and that the workspace stayed dormant. Billing support is free. If the normal menu is blocked, you can open the support‑request page directly and choose Billing there.

Was this answer helpful?

1 person found this answer helpful.
0 comments No comments

1 additional answer

Sort by: Newest
  1. AI answer

    2026-06-20T21:12:03.2333333+00:00

    Microsoft Support can investigate unexpected Azure Databricks charges and, where appropriate, apply credits or refunds. Billing support is free.

    Because the portal flow can be confusing and sometimes loops without clearly offering a ticket option, follow these steps carefully to open a billing support request for this Databricks cost issue:

    1. Open the support request creation page directly:
      • Go to: https://portal.azure.com/#create/Microsoft.Support
    2. Start a billing case:
      • In the search box, type billing and select it.
      • Choose Billing as the support area and select Next.
    3. Select the affected subscription:
      • From the subscription dropdown, select the subscription where the Databricks charges appeared.
      • Click Next.
    4. Bypass the automated “solutions” and force creation of a ticket:
      • After a short delay, several options or “solutions” will appear.
      • Select the Create a support request button at the top.
      • Do not click any of the suggested solutions.
    5. Fill in the case details:
      • Provide a clear summary such as: “Unexpected Azure Databricks charges – workspace auto-started after 6+ months dormant; request investigation and credit.”
      • Choose an appropriate Problem type and Problem subtype under Billing (for example, unexpected charges / cost dispute), even if not an exact match.
      • Click Next.
    6. Complete Additional details:
      • After another short delay, some automated solutions may appear again.
      • Click Return to support request (upper-left) instead of selecting any solution.
      • Back on the support case form, click Next to go to Additional details.
      • Provide:
        • Subscription ID.
        • Time window of the unexpected Databricks usage (for example, January 2026).
        • Confirmation that the workspace/app had been dormant for ~6 months and that no user logged in or started clusters/jobs.
        • The total unexpected charge amount and that this was a demo environment.
      • Click Next to go to Review + create, then click Create.

    After submission, an automated email is typically sent within a few minutes, and a billing engineer usually responds within one business day. They can:

    • Verify the Databricks usage and associated Azure resources.
    • Check whether serverless or other background Databricks features (for example, data quality monitoring, Predictive Optimization, materialized views, or other serverless jobs) generated the charges.
    • Determine whether a one-time courtesy credit or refund is appropriate.

    To avoid future surprises once this is resolved, consider:

    • Using budgets and alerts in Databricks and Azure Cost Management to detect unexpected spend early.
    • Using Databricks cost dashboards and billing system tables to identify which features (for example, serverless jobs, data quality monitoring, or other serverless-billed features) generated usage.

    If the portal still does not allow creating a ticket after following the steps above, repeat step 1 in a different browser or in a private/incognito window and ensure the correct subscription is selected.


    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.