Runbooks
Runbook Coverage
Summary
A paging alert, known recurring failure mode, and high-risk procedure has a documented runbook used to guide the response.
Standards
- A paging alert MUST have a corresponding runbook describing its response.
std-ops-runbook-coverage-01 - A high-risk or recurring operational procedure, such as remediating a data quality issue or executing a standby failover, MUST be captured in a runbook.
std-ops-runbook-coverage-02 - A failure mode newly identified through an incident review SHOULD receive a runbook where it is reasonably likely to recur.
std-ops-runbook-coverage-03 - A responder SHOULD use an applicable runbook to guide the response and depart from it where incident conditions require different action.
std-ops-runbook-coverage-04
Related Standards
Implements These Principles
Runbook Content
Summary
A runbook describes its steps, applicability, resolution outcome, and escalation path in enough detail for a person or an AI tool to execute.
Standards
- A runbook MUST describe the steps needed to diagnose and resolve a failure mode or carry out an operational procedure, in enough detail to execute them.
std-ops-runbook-content-01 - A runbook MUST state the condition under which it applies and the outcome confirming the failure mode's resolution or the operational procedure's completion.
std-ops-runbook-content-02 - A runbook SHOULD identify the escalation path to follow where its documented steps do not resolve the failure mode or complete the operational procedure.
std-ops-runbook-content-03
Implements These Principles
Runbook Maintenance
Summary
A runbook is version-controlled, updated with relevant service changes, and removed from active use once it no longer applies.
Standards
- A runbook MUST be version-controlled in a discoverable location.
std-ops-runbook-maintenance-01 - A runbook SHOULD be updated in the same change that materially alters the procedure, service behaviour, or failure mode it documents.
std-ops-runbook-maintenance-02 - A runbook MUST be removed from active operational documentation once it no longer applies.
std-ops-runbook-maintenance-03
Related Standards
Implements These Principles
Runbook Access
Summary
An alert or dashboard links directly to its runbook, which remains accessible to responders without source-repository access.
Standards
- A paging alert or dashboard indicator tied to a known failure mode SHOULD link directly to the runbook covering it.
std-ops-runbook-access-01 - A runbook MUST be accessible to responders without requiring direct access to its source repository.
std-ops-runbook-access-02
Related Standards
Implements These Principles
Runbook Automation
Summary
A failure response is automated where possible, and its runbook still documents how to confirm and intervene in that automation.
Standards
- A failure mode's response or an operational procedure's execution SHOULD be automated where possible.
std-ops-runbook-automation-01 - A runbook for an automated failure response or operational procedure MUST document how to confirm the automation executed correctly and how to intervene where it does not.
std-ops-runbook-automation-02
Implements These Principles
Runbook Validation
Summary
A runbook is validated before responders rely on it; gaps found during testing are corrected promptly, and gaps found during an incident are tracked afterward.
Standards
- A runbook MUST be validated before operational use through testing or review appropriate to the procedure's risk, and SHOULD be revalidated after a material change.
std-ops-runbook-validation-01 - A gap or inaccuracy found while testing a runbook SHOULD be corrected promptly.
std-ops-runbook-validation-02 - A gap or inaccuracy found while a team member or an AI tool relies on a runbook during a real incident MUST be tracked for correction after the incident.
std-ops-runbook-validation-03