Observability
Observability is the degree to which a system's internal state can be understood from its external outputs — typically metrics, logs, and traces — allowing engineers to diagnose novel, unanticipated problems without having to ship new code…
Definition
Observability is the degree to which a system's internal state can be understood from its external outputs — typically metrics, logs, and traces — allowing engineers to diagnose novel, unanticipated problems without having to ship new code to add visibility.
Overview
Traditional monitoring answers questions you already thought to ask in advance — is CPU above 90%, is the service returning errors. Observability aims at a broader goal: giving engineers enough rich, correlated data about a system that they can answer questions they didn't anticipate when they built it, which matters enormously in complex microservices systems where failure modes are often novel and specific to a particular combination of conditions. The discipline is commonly described around three foundational pillars: metrics (numeric time-series data, often visualized in Grafana dashboards fed by systems like Prometheus), log aggregation (centralized, searchable event records), and distributed tracing (the path a single request takes across services). OpenTelemetry has emerged as the standard way to instrument applications to produce all three in a vendor-neutral format, which can then flow into platforms like Datadog, New Relic, or Dynatrace. Observability underpins nearly every other practice in the site reliability engineering (SRE) toolkit: you can't set a meaningful Service Level Objective (SLO) without reliable metrics, you can't run chaos engineering experiments without a way to observe what actually happened, and effective incident management depends entirely on being able to quickly localize a novel failure. Courses like CI/CD with GitHub Actions touch on the delivery side of this pipeline, but observability itself is what makes the running system legible once code is deployed.
Key Concepts
- Built on three foundational pillars: metrics, logs, and distributed traces
- Designed to answer novel, unanticipated questions, not just pre-defined dashboards
- Standardized instrumentation increasingly via OpenTelemetry across languages
- Correlates data across pillars to speed up root cause analysis
- Underpins SLO tracking, error budgets, and effective incident response
- Distinguishes 'known unknowns' monitoring from 'unknown unknowns' investigation
Use Cases
Frequently Asked Questions
From the Blog
Understanding Monitoring and Observability
Monitoring tells you when something is wrong; observability lets you ask why. Learn how metrics, logs, and traces work together to keep modern systems healthy.
Read More Cloud & CybersecurityThe Three Pillars of Observability and How to Use Them
Metrics tell you something is wrong, traces tell you where, and logs tell you why. Learn what question each signal answers, where each one goes blind, and how to instrument a service so an incident becomes a short investigation rather than a guessing game.
Read More