feat(monitoring): Grafana alerting → Telegram for homelab

- Secret grafana-telegram: bot token + chat ID (env var injection)
- ConfigMap grafana-alerting: provisioning files for contact point,
  notification policy, and 4 alert rules
  * Pod CrashLoopBackOff (for: 1m, noData: OK)
  * Disk > 80% on non-tmpfs filesystems (for: 5m)
  * RAM > 85% (for: 5m)
  * Pod Failed/Unknown (for: 3m, noData: OK)
- Deployment: TELEGRAM_* env vars from secret + alerting volume mount

Token interpolated via ${TELEGRAM_BOT_TOKEN} in provisioning YAML.

monitoring/deployment-kube-prometheus-stack-grafana.yaml

@@ -152,6 +152,16 @@ spec:
           value: /etc/grafana/provisioning
         - name: GF_UNIFIED_STORAGE_INDEX_PATH
           value: /var/lib/grafana-search/bleve
+        - name: TELEGRAM_BOT_TOKEN
+          valueFrom:
+            secretKeyRef:
+              name: grafana-telegram
+              key: TELEGRAM_BOT_TOKEN
+        - name: TELEGRAM_CHAT_ID
+          valueFrom:
+            secretKeyRef:
+              name: grafana-telegram
+              key: TELEGRAM_CHAT_ID
         image: docker.io/grafana/grafana:12.4.2
         imagePullPolicy: IfNotPresent
         livenessProbe:
@@ -213,6 +223,8 @@ spec:
           subPath: provider.yaml
         - mountPath: /etc/grafana/provisioning/datasources
           name: sc-datasources-volume
+        - mountPath: /etc/grafana/provisioning/alerting
+          name: grafana-alerting
       dnsPolicy: ClusterFirst
       enableServiceLinks: true
       initContainers:
@@ -270,4 +282,8 @@ spec:
         name: sc-dashboard-provider
       - emptyDir: {}
         name: sc-datasources-volume
+      - configMap:
+          defaultMode: 420
+          name: grafana-alerting
+        name: grafana-alerting