Operations & Observability

Overview

Operations and observability standards require production services to expose their behaviour and support operational response and recovery.

Choose a standard below to read it in full.

A service's log output is structured, traceable, and free of sensitive data or unnecessary detail.

A service's metrics remain accurate, visible, and bounded, with automated and actionable alerts.

Trace context propagates end-to-end through accurately structured spans and deliberate sampling.

Telemetry remains consistent across services, safe to introduce, and delivered alongside the change it observes.

Telemetry is centralised, portable, and available for resilient operations.

Recovery is designed and validated against defined objectives.

A runbook documents and validates the steps to resolve known failures and perform high-risk procedures.