# Metrics, Logs and Traces — Prometheus + Grafana

Source: https://www.skillbyai.com/en/prometheus-grafana/f-signals

> Three complementary signals.

## What each signal answers

**Metrics** are numbers sampled over time (requests per second, error ratio, memory used): cheap to store, fast to query and ideal for dashboards and alerts. **Logs** are detailed event records that explain what happened in a specific case. **Traces** follow one request across services to show where time was spent. Prometheus focuses on metrics; Grafana can display all three when connected to tools such as Loki (logs) and Tempo (traces).

## Numbers over time

Prometheus pulls numeric measurements from your services and stores them as labelled time series.

![Three ideas: observability signals, architecture, the data model.](assets/figures/prometheus-grafana/section-1-map.svg) — Figure 1.1 — Signals, architecture and data model.

## A car dashboard, a logbook and a GPS track

The speedometer (metrics) tells you something is wrong at a glance, the logbook (logs) records each event in detail, and the GPS track (traces) shows the exact route a single trip took.

## Alert on metrics, debug with logs and traces

Metrics detect problems cheaply; logs and traces explain them.

**Quiz:** Which signal is best for cheap, fast alerting on error rates?

- [ ] Traces only
- [ ] Logs only
- [x] Metrics
- [ ] Screenshots

*Answer:* Metrics. Aggregated numbers are cheap to evaluate.