SolidBowtie

Critical service outage

Availability as a security property: zone failures, bad deployments, DDoS and expired certificates against a 24/7 platform.

Hazard24/7 platform running on cloud infrastructure
Top eventCritical production service unavailable

5 threats 16 barriers 3 consequences safety-critical barriers marked

Use this template free. Copy “Critical service outage” into your own workspace and adapt every barrier — no credit card.

Start free with this template →

Threats & preventive barriers

Each threat is a credible pathway to the top event; its barriers interrupt that pathway before control is lost.

Cloud provider zone or region failure

  • Multi-AZ redundancy with auto-failover (A.8.14) ★ critical

    Passive (hardware) · 90% effective · Owner: Head of Infrastructure

  • Regional DR with tested runbooks (A.5.30)

    Socio-technical · 75% effective · Owner: Head of Infrastructure

    Runbooks drift from the live architecture

    📅 Regional failover drill · semiannual keep-alive task

Faulty change or deployment

  • Change management with CAB (A.8.32)

    Socio-technical · 72% effective · Owner: Change Manager

  • CI pipeline: tests, canary, auto-rollback (A.8.31) ★ critical

    Active hardware · 85% effective · Owner: DevOps Lead

DDoS attack

  • Edge DDoS protection (A.8.20) ★ critical

    Active hardware · 88% effective · Owner: Head of Infrastructure

    Application-layer attacks under volumetric thresholds

  • Auto-scaling capacity headroom

    Active hardware · 75% effective · Owner: Cloud Platform Lead

Expired certificate or leaked secret

  • Automated certificate lifecycle (A.8.9)

    Continuous · 85% effective · Owner: DevOps Lead

    📅 Certificate expiry scan · weekly keep-alive task

  • Central secrets vault with rotation (A.8.24)

    Active hardware · 80% effective · Owner: Security Engineering Lead

Capacity exhaustion under growth

  • Capacity management & forecasting (A.8.6)

    Socio-technical · 70% effective · Owner: Cloud Platform Lead

  • Monitoring & alerting with SLOs (A.8.16)

    Continuous · 82% effective · Owner: SRE Lead

Consequences & recovery barriers

Once the top event happens, recovery barriers limit each consequence.

SLA breaches and service credits

  • Incident command & escalation process (A.5.24)

    Socio-technical · 80% effective · Owner: SRE Lead

  • Public status page & customer comms

    Behavioural (human) · 75% effective · Owner: Support Lead

Data loss for in-flight transactions

  • Durable queues & cross-region replication (A.8.14) ★ critical

    Passive (hardware) · 85% effective · Owner: Head of Infrastructure

  • Point-in-time recovery backups (A.8.13)

    Passive (hardware) · 88% effective · Owner: Backup Administrator

    📅 PITR restore validation · quarterly keep-alive task

Customer churn after repeat outages

  • Blameless post-incident reviews (A.5.27)

    Socio-technical · 78% effective · Owner: SRE Lead

  • Problem management of repeat causes (A.5.27)

    Socio-technical · 70% effective · Owner: Engineering Director

Copying this template gives you a fully editable bowtie: barrier owners arrive as role suggestions, escalation factors sit on their barriers, and keep-alive activities are scheduled from the moment you copy. New to the notation? Read what the bowtie method is or browse the practical guides.

Use this template free. Copy “Critical service outage” into your own workspace and adapt every barrier — no credit card.

Start free with this template →