Identify schedule/location to execute self managed AI upgrade tests

When Add gitlab-qa scenario to test upgrades of AI c... (gitlab-org/gitlab-qa#730 - closed) is complete we will need to find an appropriate location/schedule to run such tests that will ensure that we can detect any changes to our infrastructure that might have an adverse impact on self managed users upgrades.

We'll need to determine the set of components that are involved in this flow and identify the set of components that should trigger these tests during their respective deployments.