You cannot keep a service reliable if you cannot see how it is behaving. Monitoring tells you when something you expected to go wrong has gone wrong, while observability gives you enough data to work out why when something unexpected breaks. In practice both rest on three kinds of telemetry: metrics, logs, and traces.
You'll start with the core ideas behind observability, then learn how Prometheus collects and stores metrics and how to run it on Kubernetes alongside Grafana dashboards. The module finishes with distributed tracing, which follows a single request as it passes through many services and shows where time is being lost.
With good visibility in place, you'll be ready for the next module, SLIs, SLOs & Error Budgets, where the data you collect is turned into measurable reliability targets.
An introduction to metrics, traces, and logs, the three types of telemetry that together describe how a system is performing.
How Prometheus collects metrics from targets, and a tour of its main components: the server, Pushgateway, Alertmanager, and data storage.
A hands-on demo of deploying Prometheus and Grafana on a Kubernetes cluster with a Helm chart, and opening their dashboards.
How traces, spans, trace IDs, context propagation, and sampling work together with OpenTelemetry to follow a request across many services.