Observability
Metrics, logs, traces: how to see what's happening in production without logging into every server.
Observability
Metrics, logs, traces, Prometheus, Grafana, alerts, SLOs.
The three observability signals: the question each one answers, and when to look at which.
DEBUG/INFO/WARN/ERROR, why levels exist, structured logging in JSON, and what never belongs in a log.
How Prometheus scrapes metrics: the pull model, metric types, the /metrics endpoint, and why labels matter.
Why visualize metrics, how dashboards and panels are built, and why a good dashboard answers a question instead of showing everything at once.
How metrics turn into alerts: alert rules, routing and grouping in Alertmanager, and cutting down the noise.
How to measure reliability in numbers: SLI as a quality metric, SLO as the target, and the error budget as your license to take risks.