Databricks Community

uhuru_furuta · Thursday

As my budget was exhausted, my serverless warehouse stopped and I extended my budget, which is now 'Under Budget'.

However, my serverless warehouse still can't start as pasted. It is about 19 hours after when I extend budget.

Do I need to do any other things to restart?

Regards,

GaneshI · Thursday

Sorry to hear you're running into this — it's a frustrating situation, especially after you've already taken the right step to extend the budget. Here are a few things to check and try:

1. Check for a spending policy "lock" state Even after a budget is extended and shows "Under Budget," the serverless warehouse may remain in a policy-suspended state that doesn't automatically lift. This is a known behavior — the warehouse doesn't always auto-resume once the budget constraint is resolved.

2. Try manually restarting the warehouse

Go to SQL Warehouses in the Databricks UI
Find your serverless warehouse and click Start
If the Start button is greyed out or returns an error, note the exact error message — it will help narrow down the cause (policy block vs. a different issue)

3. Verify the budget policy is correctly re-applied

Navigate to Settings > Budget Policies
Confirm the policy associated with the warehouse is in an active/healthy state and not still flagged as over-limit
Sometimes a detach and re-attach of the budget policy to the warehouse can reset the state

4. Check account-level spending limits If you're on a Databricks trial or have an account-level spending cap (separate from the budget policy), that may also be blocking the warehouse even if the policy-level budget looks fine.

5. Open a Support Ticket Since it's been ~19 hours and the warehouse still won't start despite the budget being resolved, this likely needs Databricks Support to manually clear the suspended state on the backend. When you reach out, include:

Your workspace URL
The warehouse ID (visible in the warehouse settings URL)
The exact error message when trying to start
Timestamp of when you extended the budget

Databricks Support can check if there's a backend lock on the resource that needs to be manually released.

uhuru_furuta · Friday

Thank you @GaneshI ,

I've already tried manual restarting and I get the error below.
=====
Clusters are failing to launch. Cluster launch will be retried.

Request to create a cluster failed with an exception: INVALID_PARAMETER_VALUE: Serverless warehouse cannot start since workspace 7474651022159303 is no longer eligible for Serverless Compute. Please contact your administrator.
=====

Also I sent an email to help@databricks.com.

Best Regards,

GaneshI · Friday

Hi @uhuru_furuta,

Thanks for sharing the exact error — that makes the issue much clearer. The error:

"Serverless warehouse cannot start since workspace 7474651022159303 is no longer eligible for Serverless Compute"

This is NOT just a budget issue anymore. Extending the budget alone won't fix this. Here's what's likely happening and how to resolve it:

Root Cause

When a workspace exceeds its budget and gets suspended, Databricks can internally mark that workspace as ineligible for Serverless Compute at the platform level — this is a backend eligibility flag that is not automatically reset when the budget is extended. It requires manual intervention.

Steps to Resolve

Step 1 — Account Admin Action (Try this first)

Log in as an Account Admin (not just Workspace Admin) at accounts.azuredatabricks.net or accounts.databricks.com depending on your cloud
Go to Workspaces → your workspace → Feature Enablement
Check if Serverless Compute is still toggled ON
If it's OFF or greyed out, toggling it back ON may restore eligibility

Step 2 — Check if Serverless is disabled at the Account level

Under Account Console → Settings → Feature Enablement, confirm Serverless is enabled globally for your account
A budget breach sometimes triggers an account-level disable, not just workspace-level

Step 3 — Contact Databricks Support (Most likely required) Since the error is an internal eligibility flag, this almost certainly needs Databricks Support to flip a backend flag for your workspace. When contacting them, provide:

Detail Value

Workspace ID	7474651022159303
Error	INVALID_PARAMETER_VALUE: Serverless warehouse cannot start since workspace is no longer eligible
Context	Budget was exhausted → extended → now Under Budget, but warehouse still blocked
Time since budget extended	~19+ hours

Ask them specifically to: "Re-enable Serverless Compute eligibility for workspace 7474651022159303 which was suspended due to budget exhaustion."

Important Note

If you are on a Databricks trial (DBU trial or free tier), budget exhaustion can permanently revoke Serverless eligibility for that workspace and it may not be restorable — you may need a new workspace with a paid plan attached.

uhuru_furuta · Sunday

Hi @GaneshI ,

Thanks. I checked step 1 and 2 though I coudn't find the menu you mentioned.

I submitted a ticket to support. I'll add sentence you told me(Re-enable Serverless Compute eligibility for workspace 7474651022159303 which was suspended due to budget exhaustion.).

I am not on a Trial.

Regards,

MoJaMa · Friday

To add to what Ganesh said, it seems that you have a Databricks Team (Account Executive and Solutions Architect). I would escalate this to them while you create the Support ticket in parallel. Based on what I can see, I think you had a Trial Account which has now expired it's trial duration.