We have a small-but-growing number of automated system status checks, but they are not prominent in the UI. Some reflect critical problems (e.g. NEXUS-18706), but there's no warning whatsoever unless an admin generates a support zip or digs into the UI.
- When a health check is failing, raise a visible alarm in the admin UI
- We don't want failing system status check alerts visible to non-admin users.
This will trigger the need to schedule system status checks being run on a periodic basis, since they are currently run on demand.