operating-production-services

SLOs (Service Level Objectives) set reliability targets; error budgets quantify acceptable downtime. Define SLOs by identifying critical user journeys, measuring SLIs (Service Level Indicators) like latency and availability, then calculate error budget as the difference between 100% and your SLO target. This guides velocity decisions and prevents over-provisioning.

Blameless postmortems follow a structured format: timeline of events, root cause analysis, impact assessment, and action items. This Skill provides standardized templates that focus on system failures rather than individual mistakes, enabling teams to extract learning and prevent recurrence while maintaining psychological safety.

Incident-response playbooks document detection triggers, escalation paths, communication templates, and resolution steps for common failure scenarios. Start by cataloging your services' failure modes, then create runbooks aligned with those scenarios. This accelerates response time and reduces mean time to resolution during outages.

Yes. Error budgets translate SLO targets into concrete deployment allowances—once budget is exhausted, focus shifts to stability over new features. This framework lets teams ship faster when systems are stable and apply brakes when reliability drifts, eliminating guesswork about safe deployment rates.

SLIs are the measurable metrics underlying your SLO: request success rate, latency percentiles (p50, p99), and error rates. Collect these via monitoring tools, then compare aggregated results against your SLO threshold. Consistent measurement enables error budget calculations and reveals when services drift out of reliability targets.

Postmortems identify systemic weaknesses—missing alerts, unclear runbooks, or architectural gaps—not individual failures. Tracked action items address root causes directly. Blameless postmortems encourage honest reporting, surfacing real problems earlier and building institutional memory across your reliability practice.