Grafana turns metrics from systems such as Prometheus into dashboards that operators can scan and explore. This hands-on course guides you through the browser workflow for connecting a data source, choosing visual forms, adding reusable controls, and defining a basic integrated alert.
You will chart a Prometheus process metric, compare Time series, Gauge, and Stat panels, add a visual threshold, and build a variable-driven instance selector. You will also create an alert rule over Grafana’s TestData source, while the final challenge asks you to reconnect Prometheus and rebuild a focused CPU dashboard independently.
What You Will Learn
- Install Grafana from its signed APT repository and run it on port 8080
- Add and test a local Prometheus data source
- Create, title, save, and verify a dashboard and visualization panel
- Choose Time series, Gauge, and Stat views for different reading tasks
- Configure a threshold and arrange panels in Grafana’s grid layout
- Create a query variable from Prometheus label values and use it in a panel query
- Define and verify an integrated Grafana alert rule using TestData
- Rebuild a Prometheus-backed operational dashboard from a concise specification
Who This Course Is For
This intermediate course is for Linux administrators, DevOps learners, and operations engineers who have metrics available and want practical experience turning them into usable Grafana dashboards.
Prerequisites: Comfort with a Linux terminal, sudo, basic browser navigation, metrics and Prometheus concepts, and simple PromQL metric selectors. No previous Grafana experience is required.
Learning environment: A Linux virtual desktop with Grafana on localhost:8080, Prometheus on localhost:9090, administrator access, and Firefox. Most work is performed manually in the Grafana UI, with API calls used for verification.
Frequently Asked Questions
Do I need to build Prometheus before taking this course?
No. The lab provides a running local Prometheus service and sample metrics. You connect Grafana to it; Prometheus installation, exporter setup, and advanced PromQL are not part of this course.
Are the dashboard variables selecting different servers?
They demonstrate real variable discovery and query substitution, but the two sample instance labels point to the same lab machine through different local addresses. The course does not build a multi-host dashboard fleet.
Will I configure a production notification channel and receive an alert?
No. You create a rule named High Traffic Alert over Grafana TestData, choose an existing contact point, and verify the saved rule. Delivery to email, Slack, PagerDuty, or another external system is not tested.
