README

04 — Monitoring

Observability stack: Prometheus + Thanos for metrics, Grafana for dashboards, Alertmanager for email/Discord alerts, and a suite of custom Python exporters covering Supabase queues, slow queries, backup status, costs, and credential rotation age.

Stack runs on vps-i1 (217.154.82.162) via Docker Compose at /opt/p24-infra/monitoring/.

Key Documents

Document	Description
monitoring-stack-operations.md	Docker Compose lifecycle — start, reload, upgrade
monitoring-prometheus-grafana.md	Prometheus scrape config, retention, Thanos sidecar/query
grafana-operations.md	Dashboard management, datasource config, image renderer
monitoring-exporters-operations.md	All custom exporters — ports, env vars, metrics exposed
alert-response-runbook.md	How to respond to each alert rule
supabase-slow-query-monitoring.md	pg-stats-exporter slow-query monitoring setup

Component Map

Component	Image	Port	Purpose
Prometheus	prom/prometheus	127.0.0.1:9090	Scrapes all targets, 15d local TSDB
Thanos sidecar	quay.io/thanos/thanos	internal	Uploads 2h blocks to Wasabi S3
Thanos query	quay.io/thanos/thanos	internal	Unified PromQL over local + S3
Alertmanager	prom/alertmanager	127.0.0.1:9093	Email via Mailgun EU
Grafana	grafana/grafana	127.0.0.1:3000	Dashboards (Thanos + Supabase PostgreSQL)
queue-exporter	custom Python	:9200	Supabase queue depths
pg-stats-exporter	custom Python	:9201	pg_stat_statements slow queries
backup-exporter	custom Python	:9220	Wasabi backup status JSON
cost-exporter	custom Python	:9210	Vercel/Supabase/Wasabi billing
vercel-exporter	custom Python	internal	Vercel deployment metrics
credential-exporter	custom Python	internal	Credential rotation age
grafana-image-renderer	grafana/grafana-image-renderer	internal	PNG screenshots for daily reports

Public URLs

URL	Service
grafana.vps-i1.infra.zintegrowana.online	Grafana (Grafana login)
prometheus.vps-i1.infra.zintegrowana.online	Prometheus (basic_auth)
alertmanager.vps-i1.infra.zintegrowana.online	Alertmanager (basic_auth)
infra.zintegrowana.online	Grafana public alias

Improvement Proposals

Proposal	Description
02-loki-logs.md	Add Loki for log aggregation
05-blackbox-synthetic.md	Blackbox exporter for synthetic probes
06-consolidate-health-checks.md	Consolidate health-check endpoints
07-status-page.md	Public status page (Uptime Kuma)
10-deployment-version-dashboard.md	Deployment version tracking dashboard
11-cost-dashboard.md	Cost dashboard in Grafana
12-cert-expiry-alerts.md	TLS cert expiry alert rules

Cross-references

• README — how to act on alerts fired by Alertmanager
• README — Mailgun SMTP used for alert emails