An Apache Spark-based analytics platform optimized for Azure.
Hi @Shubham , Hope you are doing well.
That error means Unity Catalog can't resolve the managed identity behind your Access Connector in the Databricks account context — it isn't a storage permissions failure, so it happens before any ADLS access is attempted. Note your error string shows Managed Identity ID: as empty, which is correct if the connector uses a system-assigned identity.
Before anything else — do you actually need to create a metastore?
Databricks began enabling new workspaces for Unity Catalog automatically on November 9, 2023, and the docs are explicit: "You must follow the instructions in this page only if you have a workspace and don't already have a metastore in your workspace region." Only one metastore is permitted per region. In the account console, go to Catalog and check whether a metastore already exists in your workspace's region — if it does, you attach your workspace to it rather than creating a new one, and this error goes away entirely.
If you do need to create one, here's the check that matters most
Rather than repeat the checklist above, let me point you at what has actually resolved this for other people, because several customers hit this error with a fully correct configuration.
On this thread the customer confirmed same tenant, same subscription, same region, system-assigned identity enabled, metastore admin privileges, and the resource ID copied directly from the connector's resource page — and still got your exact error. The resolution from Manoj Kumar Boyini (Microsoft) was:
"In some cases, the Databricks UI can fail to resolve the Access Connector even when the configuration is correct, while the CLI/API path succeeds because it bypasses UI-level validation. When using a system-assigned Access Connector, only the access_connector_id is required (the same value you're already using in the UI)."
So try creating the storage credential via the Databricks CLI instead of the account console UI, passing only the access connector ID with no managed identity ID. If it succeeds via CLI, that confirms the problem is UI-side rather than an identity or permissions issue — and you've saved yourself from rebuilding resources that were never broken.
CLI setup: https://learn.microsoft.com/azure/databricks/dev-tools/cli/
Quick checks worth doing first (they take two minutes)
- Confirm the identity is actually provisioned. Open the Access Connector → Managed Identity tab and verify a principalId / objectId is present. If there's no principal ID, the identity isn't ready yet. If you created the connector and retried immediately, wait 5–10 minutes for propagation and try again — this is a very common cause of this specific error.
- Region alignment. The metastore region must match the Access Connector and storage account region.
- Role scope. If Storage Blob Data Contributor is only assigned at the storage account level, try assigning it at the container level as well.
- Account admin. Metastore creation happens at account level, so confirm you're an account admin in the Databricks account console — workspace admin is not sufficient.
A note on your post
You've included your full subscription ID and Databricks Account ID in the question. Neither is a secret on its own, but you may want to edit them out or mask them — the error is reproducible without them.
References
- Create a Unity Catalog metastore — https://learn.microsoft.com/azure/databricks/data-governance/unity-catalog/create-metastore
- Use Azure managed identities in Unity Catalog to access storage — https://learn.microsoft.com/azure/databricks/connect/unity-catalog/cloud-storage/azure-managed-identities
- Manage Unity Catalog metastores — https://learn.microsoft.com/azure/databricks/data-governance/unity-catalog/manage-metastore
- Databricks CLI — https://learn.microsoft.com/azure/databricks/dev-tools/cli/
Kind Regards,
Microsoft Support Team.