ERROR: Rapid fire connection error without exit

[nstans]

We have a CICD test running from a simple python script in a git action. Authenticating with a service principal pat token against a Databricks warehouse.

Since the release of the most recent version we have been experiencing extended intermittent failures.

The git action output prints out the error Service Principal does not have permission to use warehouse. This error repeats around every 20 seconds. Yesterday we were unable to cancel the git action while this connect to warehouse step was occurring and it repeated printing out that error for about two hours.

There are no loops in the script, it executes a single databricks query after initializing a single connection to a warehouse.

Once git eventually let us cancel the action we reran the action without changing any permissions or touching Databricks at all and it connected and executed as expected in under 30 seconds, no errors reported.

The issue has occurred again today but it did not prevent us cancelling the git action, and again today after restarting the git action it executed as expected.

Are there any changes we should be aware of that would allow us to prevent this happening in the future.

[msrathore-db]

Can you please share more on what version of connector you're using and what script are you running along with the logs?