Notebooks owner shows 'Unknown'

[SKakarla]

Hi All,

We are using CI/CD to deploy notebooks from GitHub and authenticating via Azure Service Principals (SPNs). Until last week, the notebook owner was correctly displayed as the SPN. However, over the past few days, the owner is now shown as "Unknown."

Despite this change, I’m still able to update and add new notebooks using the same SPNs, so functionality isn’t blocked. This issue is appearing across all environments and multiple Databricks workspaces, each using different SPNs.

This doesn’t appear to be a problem on our end, if it were limited to one workspace or a single SPN, I would suspect a configuration issue. But since it’s consistent across all Databricks instances in our tenant, I'm curious to understand what’s causing this change.

[lingareddy_Alva]

@SKakarla 

The "Unknown" owner problem appearing suddenly across multiple environments does suggest a Databricks-side change rather than something in your configuration.
This could be related to a recent update in how Databricks handles service principal authentication or displays ownership metadata. A few potential explanations:

1. There might have been a change in how Databricks resolves or displays SPN identities in the notebook metadata
2. Databricks could have modified permissions or identity handling for automated processes
3. There could be a new security update that's affecting how service principal information is displayed

Since the functionality is still working (you can still update and add notebooks), this appears to be primarily a display/metadata issue rather than an authentication problem.
To troubleshoot this, I'd recommend:

1. Checking if there are any recent Databricks platform updates or release notes that might explain this change
2. Opening a support ticket with Databricks to confirm if this is expected behavior from a recent update
3. Verifying if the actual notebook metadata (via API calls) still contains the SPN information even though the UI shows "Unknown"

[-werners-]

Same here, but indeed it does not seem to have any impact at all.
So I guess something changed in the databricks backend as suggested before.