The one-glance health check

GET /bng/health rolls the box up to a single status — ok, degraded, or fault — using a deliberate rule:

• fault — the data plane (VPP) is down, or all RADIUS endpoints are unreachable. Subscribers are affected; act now.
• degraded — exactly one of a redundant pair of RADIUS endpoints is unreachable. Still serving; investigate.
• ok — serving normally.

Notably, a full address pool is reported but does not, by itself, drive the status to fault — it is an operational condition to act on (add addresses or a node), not a service-down event. The same underlying signals are also exposed as Prometheus gauges (vpp_up, radius_up) so your Alertmanager can page on them directly. This pull-based model replaces SNMP traps.