WorkspaceClient in Databricks SDK for Python to Connect to Foreign Workspace

[adamfortuno]

Running the following code in a serverless notebook on Azure Databricks:

WORKSPACES = [
    {'workspace': 'dev', 'url': 'https://adb-123.x.azuredatabricks.net/'},
    {'workspace': 'test', 'url': 'https://adb-456.x.azuredatabricks.net/'}
]

for workspace in WORKSPACES:
    print(f"Collecting data from workspace {workspace['workspace']}")
    print(f"...workspace URL is '{workspace['url']}'")

    ## Get the Databricks name and id for workspace hosting the notebook
    wksp = WorkspaceClient(host=workspace['url'])
    workspace_id = wksp.get_workspace_id()

    print(f"...workspace identifier is '{workspace_id}'")

Despite the routine's attempt to connect to a foreign workspace. It appears to ONLY connect to the workspace the notebook is running with-in. I confirmed the workspace URLs.

Collecting data from workspace databricks-dev
...workspace URL is 'https://adb-123.3.azuredatabricks.net/'
...workspace name is 'dev' and identifier is '123'
Collecting data from workspace databricks-test
...workspace URL is 'https://adb-456.6.azuredatabricks.net/'
...workspace name is 'dev' and identifier is '123'

I skimmed the docs on the WorkspaceClient constructor. I'm not getting an authentication error. It wasn't clear to me from the docs whether subsequent instantiations of WorkspaceClient leverage my notebook's security context.