Skip to main content
OpsRoadmapsSign in
Reliability & Runtime OpsLast reviewed 2026-09-15

SRE

Engineering reliability into systems with SLOs, error budgets, and blameless incident learning.

Overview

Read the full SRE guidance
Site Reliability Engineering treats reliability as a software problem: measure it with Service Level Objectives, budget for failure with error budgets, and spend engineering time on automation instead of manual toil. When the budget is healthy, teams ship; when it burns, releases freeze until reliability recovers — a negotiated contract between product velocity and stability. OpsRoadmaps assesses SRE through SLO definition quality, error-budget policy in actual use, incident response discipline (on-call, runbooks, blameless postmortems that produce tracked actions), and toil reduction. A team that says "we do SRE" but cannot state its current error-budget burn for its top service will see the gap reflected in its score, not papered over.

Capabilities

  • Alerting

    Symptom-based, SLO-backed alerts that page a human only when a user-visible promise is burning.

  • Incident Response

    Detecting, coordinating, and resolving unplanned interruptions with defined roles and severities.

  • SLO Management

    Defining, measuring, and governing Service Level Objectives and their error budgets.

Practices

  • Blameless Postmortems

    Incident reviews that fix systems, not people — with tracked action items.

  • Error Budget Policy

    A written contract trading release velocity for reliability when budgets burn.

  • Toil Budgeting

    Measuring and capping manual operational work to fund automation.

Metrics

  • Availability

    Achieved uptime against the defined SLO window.

    Unit: %. Good direction: up.

  • Error Budget Burn Rate

    Speed at which the error budget is being consumed.

    Unit: × budget/hour. Good direction: down.

  • MTTR — Mean Time to Restore

    Average time from incident start to service restoration.

    Unit: hours. Good direction: down.

Architecture patterns

Reference architectures and their trade-offs live in the blueprints library.

Maturity

Maturity for sre is measured, not guessed — every score traces to your answers. See how maturity is scored.

Put it to work

Assess your SRE maturity

Turn gaps into a roadmap

Generate a prioritized, effort-aware improvement plan from your results.

Grab the operating checklists

Repeatable checklists for the ceremonies this discipline runs.

Browse the runbooks

Step-by-step responses for when this area catches fire.

Related Ops disciplines

Sources

SRE — OpsRoadmaps | OpsRoadmaps