Outage Intelligence & Emergency Response Foundations

Glossary

Blameless Postmortem

A structured incident review that assumes good faith and competence, directing corrective effort at systems and processes rather than individuals, so that people disclose what truly happened and the organization can fix underlying causes.

Blast Radius

The extent of damage a single failure or compromise can cause; shrinking it through network segmentation and least-privilege access limits how far one bad event — accidental or adversarial — can spread.

Common-Mode Failure

A single event that disables both a primary system and its supposed backup at once because they share a hidden dependency — the same grid, fuel supply, location, network, or credentials — defeating redundancy that looked sound on paper.

Failover

The automated handoff of workload from a failed system or site to a healthy standby, typically across independent locations (multi-region redundancy); its reliability is only proven by realistic testing, not by its presence in a plan.

Recovery Point Objective (RPO)

The maximum amount of data, measured as a span of time, that an organization can afford to lose in a disaster; it governs how frequently data must be protected and, like recovery time, is a business decision about acceptable loss.

Recovery Time Objective (RTO)

The maximum length of time a system can be unavailable before the harm becomes unacceptable; a business decision that trades cost against tolerable downtime, since tighter targets require more expensive architecture.

Root-Cause Analysis

The practice of asking 'why' repeatedly past the first plausible explanation until reaching an underlying condition — a design assumption, process, or missing safeguard — that the organization can actually change, rather than stopping at a person or a single failed part.

Time-to-Detect / Time-to-Recover

Two distinct intervals within an incident: the gap between a failure starting and anyone noticing (detect), and the gap between noticing and restoring service (recover); a long time-to-detect often means customers knew before the operator did.