What is a runbook?
Also called: playbook
Runbook: A written procedure for handling alerts, detailing meaning, confirmation, mitigation steps, and escalation paths.
A runbook is a written procedure for handling a specific alert or operational task. It covers what the alert means, how to confirm it, the steps to fix or mitigate, and when to escalate. Every alert that can page a person should link to a runbook. An alert without one asks the responder to improvise at 03:00.
Good runbooks are short and contain the exact commands. They are updated after every incident where they were wrong. Steps that are run the same way every time are candidates for automation. This leaves the runbook to explain when to run it.