Not able to create metastore in azure databricks premium 14 day free trail from manage account

Question

Not able to create metastore in azure databricks premium 14 day free trail from manage account

data engineerlin 0

Hello,

I am not able to create a metastore in Azure Databricks Premium 14-day free trial from the Manage account

I have tried both the UI and databrick cli and getting contously the below error.

UI error:

Azure Managed Identity Credential with Access Connector Id /subscriptions/---/resourceGroups/---/providers/Microsoft.Databricks/accessConnectors/-- and Managed Identity ID: for Account Id: --- could not be found.

CLI error:

< HTTP/2.0 404 Not Found (Error: Not Found) pid= sdk=true

17:09:46 Debug: non-retriable error: Not Found pid= sdk=true

Error: Not Found

17:09:46 Info: failed execution pid= exit_code=1 error="Not Found"

17:09:46 Debug: no telemetry logs to upload pid=

Note: Azure Databricks Access Connector is in the same region as the workspace, has the correct system/user-assigned identity enabled, and that identity has “Storage Blob Data Contributor” on the storage account.

I have been stuck here for a long time. Please help me out.

Himaja Y 375 Reputation points Microsoft External Staff Moderator

2026-08-21T12:34:18.6766667+00:00
Hi @data engineerlin ,

Thanks for sharing the details.

Based on the error, the issue appears to be that Azure Databricks is unable to recognize or resolve the Access Connector in the Databricks account. Since the Storage Blob Data Contributor permission is already configured, this is less likely to be an RBAC issue.

The key error is:

Azure Managed Identity Credential with Access Connector Id ... could not be found for Account Id ...

We recommend checking the following:

Confirm Account Admin access

Go to Databricks Account Console → Settings → Identity and access**.

Make sure your user is an Account Admin.

Azure Subscription Owner/Contributor permissions alone are not sufficient for creating a metastore.

Verify the Access Connector

Open the Access Connector in the Azure portal.

Make sure the managed identity is enabled.

Copy the Access Connector Resource ID again and make sure the same ID is being used while creating the metastore.

Verify storage permissions

On the ADLS Gen2 storage account, confirm that the Access Connector's managed identity has: Storage Blob Data Contributor

If the role was assigned recently, allow some time for RBAC propagation.

Check the Databricks account and tenant

Make sure the workspace and Access Connector are associated with the expected Microsoft Entra tenant.

Also confirm that the metastore is being created from the same Databricks account that owns the workspace.

Recreate the Access Connector Since this is a trial/POC environment, we recommend recreating only the Access Connector rather than recreating the entire Databricks workspace.

Delete the existing Access Connector.

Create a new Access Connector for Azure Databricks.

Enable the managed identity.

Assign Storage Blob Data Contributor to the new identity on the storage account.

Wait a few minutes for the permission to propagate.

Use the new Access Connector Resource ID when creating the metastore.

The CLI 404 Not Found together with the UI error suggests that Databricks is unable to find or resolve the Access Connector in the current account context.

If the issue continues after recreating the Access Connector, collect the Databricks Account ID, Workspace ID, Access Connector Resource ID, region, and exact timestamp, and raise a Microsoft Support request so the Databricks backend account/resource association can be checked.
data engineerlin 0 Reputation points

2026-08-22T05:49:46.8466667+00:00

Thank you for your quick response.

2 answers

Your answer

data engineerlin 0 Reputation points

2026-08-22T05:49:46.8466667+00:00

Thank you for your quick response.

Answer 1

Hi @data engineerlin ,

Thank you for sharing the resolution.

Based on the final findings, the issue was not related to the Azure Databricks Access Connector, managed identity permissions, storage RBAC, or the storage account configuration.

Root cause:

The Databricks AccountAdmin CLI profile was configured with the workspace URL instead of the Databricks Account Console URL (https://accounts.azuredatabricks.net).

As a result:

Account-level Unity Catalog and Metastore API calls were sent to the wrong endpoint.
The Databricks CLI returned 404 Not Found errors.
The Catalog section in the Account Console displayed "Not Found".
Metastore creation and management operations failed even though the Access Connector and storage permissions were configured correctly.

Resolution:

Updated the AccountAdmin CLI profile to use the Databricks Account Console host (https://accounts.azuredatabricks.net).
Created a separate CLI profile for the workspace URL.
Re-authenticated both profiles.
Verified that account-level commands could successfully list and manage metastores.
Recreated and assigned a new Unity Catalog metastore to the workspace.
Recreated the Access Connector and storage credential after the previous configuration had been removed.
Confirmed the new metastore appeared correctly in both the Account Console and Workspace Catalog UI.

I'm glad to hear the issue has been resolved and Unity Catalog is now functioning as expected. Thanks for sharing the root cause and resolution with the community. Thank you for sharing the resolution.

Based on the final findings, the issue was not related to the Azure Databricks Access Connector, managed identity permissions, storage RBAC, or the storage account configuration.

Answer 2

The issue is fix now.

Unity Catalog Metastore Setup Fix

Initial problem: Account-level Databricks CLI commands (databricks account metastores ...) kept failing with Error: Not Found, and the account console's Catalog section also returned "Not Found."

Root cause discovered: The AccountAdmin CLI profile was pointing to the wrong host. Running databricks auth login --host <workspace-url> earlier without a --profile flag had overwritten the AccountAdmin profile's host with the workspace URL instead of the account console URL (https://accounts.azuredatabricks.net). Every account-level API call was silently hitting the wrong endpoint as a result.

Complication along the way: While troubleshooting, an existing (auto-provisioned) Access Connector for Azure Databricks was deleted, which broke the storage credential tied to the workspace's original metastore — leaving the workspace's Catalog page showing "Metastore does not exist."

Steps taken to fix it**:**

Fixed the CLI profile — edited the local Databricks config file directly to correct the AccountAdmin profile's host to the account console URL, and added a separate profile pointing to the actual workspace URL. Re-authenticated both.

Confirmed account-level access worked — listing metastores now returned real results instead of errors.

Discovered the old metastore no longer existed at the account level, so it couldn't be deleted or repaired directly.

Created a new metastore pointing to the correct cloud storage location.

Reassigned the workspace to the new metastore using the assignment "create" command (which overwrites any existing assignment).

Recreated the Access Connector for Azure Databricks, and granted it the correct storage access role on the storage account.

Recreated the storage credential, correcting a JSON formatting issue along the way (fields needed to be at the top level of the request body, not nested under an extra wrapper key).

Verified — the Catalog page in the workspace UI cleared, and the new metastore showed up correctly in the account console.

End result: a fully working Unity Catalog metastore, correctly authenticated and assigned to the workspace.

CLI Commands Used to Fix the Issue

1. Check authentication profiles

bash

databricks

2. Fix the account-level profile (edit config file directly)

bash

nano

Corrected the AccountAdmin block to use the account console host, and added a separate profile block for the workspace host.

3. Re-authenticate both profiles

bash

databricks

4. Verify account-level access now works

bash

databricks

5. Check the broken/stale metastore assignment

bash

databricks

6. Create a new metastore

bash

databricks

7. Reassign the workspace to the new metastore (overwrites old assignment)

bash

databricks

8. Verify the new assignment

bash

databricks

9. Create the storage credential (fields at top level, no wrapper key)

bash

databricks

That's the full sequence, generalized with placeholders in place of your actual IDs/names.