Monitoring

Service Endpoints

Service	Port	Domain
API	3000	api.moenet.work
Bot	8443	bot.moenet.work
Prometheus	9090	prom.moenet.work
Grafana	3002	grafana.moenet.work

Health Checks

API Health

curl https://api.moenet.work/health
# {"status": "ok"}

Bot Health

# Check webhook status
curl https://api.telegram.org/bot<token>/getWebhookInfo

Agent Health

Each agent exposes a local HTTP endpoint:

curl http://localhost:24368/health

Prometheus Metrics

API Metrics

curl https://api.moenet.work/metrics

Exposed metrics include:

• http_requests_total — Request count by endpoint and status
• http_request_duration_seconds — Request latency histogram
• active_sessions_total — BGP session count by status
• agent_heartbeat_timestamp — Last heartbeat per node

Agent Metrics

The agent reports metrics to the Control Plane every 60s:

• Per-session RTT (ping latency)
• Per-session traffic (rx/tx bytes)
• Per-session route count (imported/exported)
• System uptime

Docker Logs

# All services
docker compose logs -f

# Specific service
docker compose logs -f api
docker compose logs -f bot
docker compose logs -f postgres

Common Alerts

Condition	Check
Agent offline	No heartbeat for > 5 minutes
Session stuck	Status QUEUED_FOR_SETUP for > 10 minutes
High error rate	> 5% of API requests returning 5xx
Database connection	PostgreSQL health check failing
Redis connection	Redis PING failing