when using the monitoring plugin with the console dashboards plugin, if a custom datasource defined in a dashboard is not found, the default in cluster prometheus is used to fetch data. This creates a false assumption to the user that the custom dashboard is working when in reality, it should fail.

How to reproduce:

• In OpenShift 4.16
• Install COO
• Enable the console dashboards plugin as documented here
• Create a dashboard that uses custom datasources as documented here. Do not create a datasource so the bug can be reproduced
• Go to monitoring -> dashboards and select the dashboard created above

Expected result

The dashboard should display an error as the custom datasource was not found