Welcome to Observability 101, a hands-on course about turning production behavior into numbers you can query, graph, alert on, and trust.

Is This Course for You?

• Is observability a new concept for you?
• Have you heard of observability before, but not quite sure where to start?
• Do you have metrics, but not sure what to do with them?
• Is your monitoring setup unreliable?

If the answer to any of these questions is yes, then this course is for you.

The course aims to teach you the fundamentals of observability and fill in the gaps people often have when starting out.

The course starts with the big picture: telemetry, monitoring, observability, metrics, logs, traces, and profiles. After that, it narrows to metrics. Metrics are not the whole observability story, but they are usually the highest-value first signal: cheap to collect, compact to store, fast to query, and good enough to power most dashboards, alerts, and SLOs.

Prerequisites

Before starting, you should be comfortable with:

• Command line: Running shell commands, using curl, and reading JSON/YAML.
• HTTP basics: Routes, methods, status codes, and services talking over HTTP.
• Basic math: Ratios, percentages, averages, and percentiles.

No Prometheus, VictoriaMetrics, OpenTelemetry, Kubernetes, or SRE background is assumed.

What You'll Learn

By the end of this course, you'll be able to:

• Explain telemetry signals and choose when metrics, logs, traces, or profiles fit a question.
• Understand what time series are and how they relate to metrics.
• Read and design Prometheus-style metric names, labels, samples, and time series.
• Choose counter, gauge, histogram, or summary based on what is being measured.
• Collect metrics from various sources.
• Query metrics with selectors, rates, aggregations, ratios, and histogram percentiles.
• Build dashboards that answer operational questions instead of dumping graphs.
• Instrument an application with safe metric names and labels.
• Create alerts and basic reliability targets from metrics.
• Protect a metrics pipeline with cardinality limits, relabeling, dropping, retention, downsampling, and deduplication.

Course Structure

The course is organized into nine modules that build on each other, followed by a capstone:

Capstone: Apply What You've Learned

A single hands-on scenario that ties everything together: metrics-driven incident response, where you use what you've learned to investigate and resolve a production issue end to end. Coming soon, once the modules it draws on are in place.

Ready to start?