Setup & Configuration

Concepts are in the System Monitoring overview. This is a how-to.

Installation

The monitoring stack is installed automatically as part of infrastructure deployment — no manual steps. That includes the single-node install (roles/infra/run.sh via openg2p-single-node.sh) and the multi-node production install (openg2p-prod). The infra Helmfile creates the observability namespace and deploys the OTel collectors, Loki, its MinIO, the Grafana Loki datasource, the alert rules and the ready-made dashboard.

See the Deployment documentation for the overall install.

Configuration

Settings are provided in prod-config.yaml (multi-node) or single-node-config.yaml (sandbox) before install (all have sensible defaults). The main ones:

Key	Default	Purpose
loki_retention_hours	168	Log retention (168h = 7 days)
loki_minio_root_user	loki	Loki object-store user
loki_minio_root_password	(auto-generated)	Loki object-store password
loki_minio_size	50Gi	Disk for Loki's MinIO
alert_smtp_*	(blank)	Email alert channel (host, from, user, password, to)
alert_slack_webhook_url	(blank)	Slack / Mattermost / Rocket.Chat webhook
alert_telegram_bot_token / _chat_id	(blank)	Telegram channel
ai_enabled	false	Optional AI layer (off by default)

Alert channels are inert until you fill in their credentials — see the Operations Guide → Alerting.

Accessing Grafana

Grafana is reached through the Rancher UI → cluster → Monitoring → Grafana (it is not exposed on its own hostname). You are signed in automatically via your Rancher session as an Editor — no separate Grafana login. That link should open Grafana Home directly.

If Grafana shows Page not found on first load, click Home in the Grafana bar (or hard-refresh). Do not add /k8s/clusters/local to the URL — for this stack the Rancher proxy path is …/api/v1/namespaces/cattle-monitoring-system/services/http:rancher-monitoring-grafana:80/proxy/.