What the people running it need.
WHAT THEY NEED TO KNOW
How to deploy it How to roll it back How to restore from backup What it depends on What normal looks like What to do when specific things fail Who to contact
WHAT A RUNBOOK IS
Instructions for a specific operational task or failure.
WHAT MAKES ONE USEFUL
Written for someone under pressure, at night, who did not build the system.
WHAT THAT IMPLIES
Exact commands No assumed knowledge Clear decision points What to do if a step fails
WHAT TO INCLUDE FOR EACH ALERT
What it means What to check What to do When to escalate
WHY
An alert with no documented response is answered by improvisation.
WHERE TO KEEP IT
Somewhere reachable when the system is unavailable.
WHY
Documentation stored only inside the failing system is useless during an outage.
WHAT TO TEST
The recovery procedures, by performing them.
WHAT TO UPDATE
After every incident.
