Skip to content

ALL-OBSE1002 · operational_risk_coverage

Judge whether telemetry answers each declared operational risk.

This is a contextual rule for all languages. Read its implementation.

Compare critical paths, failure modes, diagnostic questions, signals, and ownership. A risk is actionable only when the retained signals can answer the questions needed to distinguish its likely failure modes without unsafe or indiscriminate collection.

Findings cite each risk, its critical path, failure modes, diagnostic questions, signals, and owner rather than a provider-supplied actionability verdict.

A signal may link to richer correlated context elsewhere. Low-risk paths may need fewer signals when the reason is explicit and the remaining questions still have reliable answers.

A payment risk with latency, outcome, dependency, and correlation signals that answer its timeout and rejection questions is actionable. A risk naming several failures but no signal that distinguishes them is gaps.

  • Cites “Observability Engineering”, debugging from unknowns
  • Cites “Site Reliability Engineering”, monitoring distributed systems
  • Cites “OpenTelemetry documentation”