What should I monitor in production for a small web app?
poze 2 months ago622 gadeen
My app is live with a few hundred users. I only find out about problems when someone messages me. What's a minimal but useful monitoring setup?
Dyagnostik IA
Se IA ki fè l. Se pa yon repons — kominote a anba a konfime oswa korije l. Toujou verifye anvan ou fye l.
1 Repons
A minimal setup that catches most problems:
1. Uptime check from outside (UptimeRobot, Better Stack, a cron + curl) on a /healthz endpoint, with alerts to your phone or Telegram. 2. Error tracking with Sentry (or GlitchTip, self-hosted): stack traces, affected users, releases. Frontend *and* backend. 3. Structured logs (JSON, e.g. pino) with a request id, sent somewhere searchable (Better Stack, Grafana Loki, Axiom). 4. The four golden signals per endpoint: latency (p95), traffic, error rate, saturation (CPU, memory, DB connections, disk). 5. Database: slow query log (pg_stat_statements), disk space, backup success. 6. Alert on symptoms, not causes: "error rate > 2% for 5 min" or "p95 > 2s", not "CPU at 80%".
OpenTelemetry is the standard if you later want traces across services.
Konekte pou di lòt moun si li te mache.
Sign in and verify your email to post an answer.