Grafana is a dashboard tool for querying and visualizing time series and Metrics. It is commonly paired with Prometheus.

It is one of the core tools used for Observability.

What people use it for

Grafana is often used to build dashboards that show:

  • CPU and memory usage
  • Request rate
  • Error rate
  • Latency percentiles
  • Queue depth
  • Service health by environment or cluster

Common pieces

  • Dashboards group related panels
  • Panels display a chart, table, stat, or heatmap
  • Data sources provide the metrics or logs
  • Variables let you switch between services, environments, or clusters
  • Alerts notify you when a condition crosses a threshold

Example queries

If Prometheus is your data source, a Grafana panel might use a query like:

rate(http_requests_total[5m])

To show request latency:

histogram_quantile(0.95, sum(rate(http_request_duration_seconds_bucket[5m])) by (le))

To show error rate:

sum(rate(http_requests_total{status=~"5.."}[5m]))

Why people use it

Grafana is useful when you want one place to explore metrics visually, compare services side by side, and make trends easy to spot.

Compare

Grafana does not usually collect metrics itself. It visualizes data from systems like Prometheus and helps make Observability easier to use.