Tune homelab availability alerts

monitoring/prometheus/alerts.yml

@@ -1,9 +1,18 @@
 groups:
   - name: homelab-availability
     rules:
+      - alert: HomelabExternalConnectivityDown
+        expr: sum(probe_success{job="blackbox-http"} == 0) >= 5
+        for: 8m
+        labels:
+          severity: warning
+        annotations:
+          summary: "External connectivity appears down"
+          description: "At least 5 public homelab endpoints are unreachable. Likely WAN, DNS, or provider issue."
+
       - alert: HomelabEndpointDown
-        expr: probe_success{job="blackbox-http"} == 0
-        for: 2m
+        expr: (probe_success{job="blackbox-http"} == 0) unless on() (sum(probe_success{job="blackbox-http"} == 0) >= 5)
+        for: 8m
         labels:
           severity: critical
         annotations:
@@ -40,10 +49,10 @@ groups:
           description: "Host memory usage is above 90%."
 
       - alert: HomelabTraefik5xx
-        expr: sum(rate(traefik_service_requests_total{code=~"5.."}[5m])) by (service) > 0
+        expr: sum(increase(traefik_service_requests_total{code=~"5.."}[5m])) by (service) >= 5
         for: 2m
         labels:
           severity: warning
         annotations:
           summary: "Traefik 5xx responses for {{ $labels.service }}"
-          description: "Traefik reports 5xx responses for {{ $labels.service }}."
+          description: "Traefik reports at least 5 5xx responses for {{ $labels.service }} within 5 minutes."