M26: harden clean observability acceptance

backend/app/api/routers/observability.py

@@ -10,7 +10,7 @@ from sqlalchemy import func, select
 from app.core.config import get_settings
 from app.core.db import SessionLocal
 from app.core.observability import OUTBOX_EVENTS
-from app.models.outbox import DEMO_SCENARIO_ERROR_CODE, OutboxEvent
+from app.models.outbox import DELIVERY_STATUSES, DEMO_SCENARIO_ERROR_CODE, OutboxEvent
 
 router = APIRouter(tags=["observability"])
 settings = get_settings()
@@ -26,8 +26,14 @@ def _refresh_database_metrics() -> None:
             ).group_by(OutboxEvent.delivery_status, "demo")
         ).all()
     OUTBOX_EVENTS.clear()
+    # Keep every time series present even when a state currently contains no rows.
+    # Stable zero-valued series make dashboards and alerts deterministic after resets,
+    # restores and fresh installations instead of turning "zero" into "no data".
+    for scenario in ("synthetic", "operational"):
+        for status in DELIVERY_STATUSES:
+            OUTBOX_EVENTS.labels(scenario, status).set(0)
     for status, is_demo, count in rows:
-        OUTBOX_EVENTS.labels(str(status), "synthetic" if is_demo else "operational").set(count)
+        OUTBOX_EVENTS.labels("synthetic" if is_demo else "operational", str(status)).set(count)
 
 
 @router.get("/metrics", include_in_schema=False)

backend/app/core/observability.py

@@ -30,7 +30,7 @@ HTTP_IN_PROGRESS = Gauge(
 OUTBOX_EVENTS = Gauge(
     "mobilityops_outbox_events",
     "Persisted outbox events by state and scenario type.",
-    ("status", "scenario"),
+    ("scenario", "status"),
 )
 DATABASE_READY = Gauge(
     "mobilityops_database_ready",

backend/app/main.py

@@ -148,11 +148,11 @@ def readiness() -> JSONResponse:
             db.execute(text("SELECT 1"))
     except Exception:  # noqa: BLE001 -- readiness must convert infrastructure errors to 503
         DATABASE_READY.set(0)
+        return JSONResponse(
+            status_code=503,
+            content={"status": "not_ready", "service": "mobilityops-api", "database": "down"},
+        )
     DATABASE_READY.set(1)
-    return JSONResponse(
-        status_code=503,
-        content={"status": "not_ready", "service": "mobilityops-api", "database": "down"},
-    )
     return JSONResponse(content={"status": "ready", "service": "mobilityops-api", "database": "up"})