Tech Terms Explained Open in the app →

Observability

DevOps · Intermediate · 4 min read

What is it?

Observability is the ability to understand what's happening inside a system from the outside, using its logs, metrics, and traces.

Explain like I'm 5

Observability is like a car's dashboard and warning lights: you can't see the engine directly, but the gauges tell you whether something's wrong and roughly where.

Why was it created?

As systems grew complex and distributed, simply checking 'is it up?' wasn't enough. Observability emerged to answer why a system behaves as it does.

Where is it used?

  • Production monitoring
  • Debugging incidents
  • Performance analysis
  • Understanding distributed systems

Why should developers care?

When production breaks, observability is how you find out why. It's a core skill for running modern services.

How does it work?

Systems emit three main signals: logs (event records), metrics (numeric measurements over time), and traces (the path of a request across services). Tools collect and correlate these so you can ask new questions about behavior.

Real-world example

When checkout slows down, engineers use traces to see which service is slow, metrics to confirm the spike, and logs to find the specific error.

Common use cases

  • Incident investigation
  • Spotting performance regressions
  • Tracing requests across services
  • Alerting on problems

Advantages

  • Faster root-cause analysis
  • Understand complex systems
  • Catch issues before users do
  • Answer unforeseen questions

Disadvantages

  • Data volume and storage costs
  • Instrumentation effort
  • Too many alerts cause fatigue

When should you use it?

For any production system, especially distributed ones, where you must diagnose issues quickly.

When should you avoid it?

Rarely — even small systems benefit; extremely simple scripts may not need full tooling.

Alternatives

Basic uptime monitoring onlyManual log inspection

Related terms

LoggingMetricsTracingOpenTelemetry

Interview questions

Beginner

  • What are the three pillars of observability?
  • How is it different from monitoring?

Intermediate

  • What is a distributed trace?
  • What's the difference between logs and metrics?

Senior

  • How do you reduce alert fatigue while keeping coverage?
  • How would you instrument a service for observability?

Common misconceptions

  • "Observability is the same as monitoring" — monitoring watches known signals; observability lets you ask new questions about unknown problems.
  • "More dashboards means more observability" — quality and correlation matter more than dashboard count.

Fun facts

  • The three pillars are logs, metrics, and traces.
  • The term comes from control theory, about inferring internal state from outputs.

Timeline

  • 2010s — Observability gains traction with microservices and cloud

Learning resources

Quick summary

Observability is understanding a system's internal behavior from its logs, metrics, and traces, so you can diagnose problems and answer new questions.

Cheat sheet

  • Three pillars: logs, metrics, traces
  • Understand behavior from outside
  • More than uptime monitoring
  • Key for debugging distributed systems

If you remember only one thing

Observability lets you understand and debug a system from its logs, metrics, and traces — not just whether it's up.