The runbook habits that made 3 a.m. survivable, and the culture that made them stick.
The dread before an on-call shift is information. It tells you the system is not ready to be operated by a tired human. We spent a quarter turning that dread into something closer to calm, and most of the work was not technical.
Dread comes from uncertainty
What makes on-call frightening is not the work — it is not knowing whether you will be able to do it. Will the alert make sense? Will there be a runbook? Will I make it worse? Every "I don't know" is a source of dread, and every one of them is fixable.
The habits that helped
- Every alert links to a runbook. No page without a documented first step.
- Runbooks are tested in game days, so they are trusted, not hoped-for.
- Escalation is encouraged, not treated as failure. Waking a teammate is a valid move.
Culture makes it stick
Habits decay without a culture that values them. When the team treats a missing runbook as a bug to fix rather than a fact of life, the safety net stays intact. When escalation is celebrated instead of judged, people use it before things get bad.
The result
The pages did not disappear, but the dread mostly did. A shift became a known quantity: clear alerts, trusted runbooks, and a team you were allowed to lean on. That is what on-call feels like when the system is built for the human holding it.