Why should you care?

Users report slowness before you notice unless dashboards watch error rate and latency.

See it live — copy this example

Paste into your MeanVerse project (Angular + Express + MongoDB), then run with ng serve / node / mongosh as noted.

import promClient from 'prom-client';

const httpDuration = new promClient.Histogram({
  name: 'http_request_duration_seconds',
  help: 'API latency',
  labelNames: ['method', 'route', 'status']
});

app.use((req, res, next) => {
  const end = httpDuration.startTimer();
  res.on('finish', () => end({ method: req.method, route: req.route?.path ?? req.path, status: res.statusCode }));
  next();
});

app.get('/metrics', async (_req, res) => {
  res.set('Content-Type', promClient.register.contentType);
  res.end(await promClient.register.metrics());
});

What happened?

• Histogram records latency per route.
• Prometheus scrapes /metrics.
• Grafana alerts when p95 exceeds SLO.

Practice next

1. Expose /metrics on internal port only.
2. Track golden signals: latency, traffic, errors, saturation.
3. Add Mongo and Redis exporter metrics.
4. Add OpenTelemetry trace from Angular to Mongo.
5. Synthetic check pings /health every 60s.

Remember

Metrics quantify user-visible pain. Prometheus + Grafana common stack. Alert on SLO burn not every blip.

Payment spike

Error rate jumps 5% during marketing email.

Outcome: Alert fires; on-call scales API pods before outage.