Monitoring Stack

Observability-Plattform für NeXifyAI. Prometheus + Grafana + Alerting.

Key Facts

• Prometheus: Port 9090
• Grafana: Port 3030 (nicht 3000 wegen WhatsApp-Bridge), admin.nexifyai.cloud/grafana
• Compose: /opt/nexifyai/monitoring/docker-compose.yml
• Grafana Auth: admin/admin
• Network: network_mode: host, GF_SERVER_HTTP_ADDR=127.0.0.1

Metriken

• 9 Metriken vom Local-AI-Worker
• 5 Alerting-Regeln
• Hostinger VPS Metrics (via node_exporter textfile)
• Intelligence API v3 (23 Quellen)

Alerts

• CPU > 80%
• RAM > 85%
• Disk > 90%
• Service-Down
• SSL-Expiry < 7 Tage

Beziehungen

• [[traefik]] — Grafana-Routing via /grafana
• [[9router]] — API-Call-Metriken
• [[hermes-agent]] — Cron-Health-Checks

Quellen

• raw/articles/agents-root.md