feat(observability): close native SigNoz backup restore canary

ops/monitoring/alerts-unified.yml

@@ -716,21 +716,23 @@ groups:
           description: "OpenClaw (192.168.0.188:8088) 已離線超過 2 分鐘"
 
       - alert: SignOzDown
-        expr: probe_success{job="blackbox-http", instance=~".*3301.*"} == 0
+        expr: >-
+          (probe_success{job="blackbox-http", instance="http://192.168.0.110:8080/api/v1/health"} == 0)
+          or absent(probe_success{job="blackbox-http", instance="http://192.168.0.110:8080/api/v1/health"})
         for: 2m
         labels:
           severity: warning
-          layer: docker-188
+          layer: docker-110
           component: signoz
-          host: "188"
+          host: "110"
           team: ops
           auto_repair: "true"
           mcp_provider: "ssh_host"
-          target_host: "192.168.0.188"
+          target_host: "192.168.0.110"
           alert_category: "devops_tool"
         annotations:
           summary: "SignOz 服務離線"
-          description: "SignOz (192.168.0.188:3301) 已離線超過 2 分鐘"
+          description: "SignOz (http://192.168.0.110:8080/api/v1/health) 已離線或監控序列缺失超過 2 分鐘"
 
       # ---- 110 Docker 層 ----
       - alert: SentryDown

ops/monitoring/alerts.yml

@@ -400,21 +400,23 @@ groups:
           description: "OpenClaw (192.168.0.188:8088) 已離線超過 2 分鐘"
 
       - alert: SignOzDown
-        expr: probe_success{job="blackbox-http", instance=~".*3301.*"} == 0
+        expr: >-
+          (probe_success{job="blackbox-http", instance="http://192.168.0.110:8080/api/v1/health"} == 0)
+          or absent(probe_success{job="blackbox-http", instance="http://192.168.0.110:8080/api/v1/health"})
         for: 2m
         labels:
           severity: warning
-          layer: docker-188
+          layer: docker-110
           component: signoz
-          host: "188"
+          host: "110"
           team: ops
           auto_repair: "true"
           mcp_provider: "ssh_host"
-          target_host: "192.168.0.188"
+          target_host: "192.168.0.110"
           alert_category: "devops_tool"
         annotations:
           summary: "SignOz 服務離線"
-          description: "SignOz (192.168.0.188:3301) 已離線超過 2 分鐘"
+          description: "SignOz (http://192.168.0.110:8080/api/v1/health) 已離線或監控序列缺失超過 2 分鐘"
 
       # ---- 110 Docker 層 ----
       - alert: SentryDown

ops/monitoring/generate_monitoring.py

@@ -54,7 +54,10 @@ def generate_prometheus_scrape_configs(registry: dict) -> list[dict]:
 
     # K8s 服務
     for svc in registry.get("services", []):
-        if svc.get("monitoring", {}).get("prometheus"):
+        monitoring = svc.get("monitoring", {})
+        if monitoring.get("prometheus") and monitoring.get(
+            "prometheus_scrape", True
+        ):
             if svc.get("type") == "k8s-deployment":
                 # K8s ServiceMonitor 風格
                 config = {

ops/monitoring/service-registry.yaml

@@ -335,6 +335,8 @@ services:
     health_type: grpc
     monitoring:
       prometheus: true
+      prometheus_scrape: false
+      external_verifier: scripts/reboot-recovery/verify-signoz-otel-grpc-runtime.sh
       sentry: false
     alerts:
       - service_down
@@ -345,11 +347,13 @@ services:
   # --- SignOz UI ---
   - name: signoz-ui
     type: docker
-    host: 192.168.0.188
-    port: 3301
-    health_endpoint: /
+    host: 192.168.0.110
+    port: 8080
+    health_endpoint: /api/v1/health
+    health_type: http
     monitoring:
       prometheus: true
+      prometheus_scrape: false
       sentry: false
     alerts:
       - service_down