Developer On-call Survival Kit - First 5 Minutes, Triage Decision Tree & 10 Production Runbooks
Most developers join an on-call rotation without a manual. The first night teaches the hard way: debugging instead of stopping the damage, rolling back into a data problem, saying nothing in the channel for 40 minutes.
This kit is the manual — written from the triage and on-call seat.
What you get
- Before your first shift - access to test, what to know, a 30-minute practice run.
- The first 5 minutes - one-page card: is it real, how big, what changed, decide, say something.
- Triage decision tree - from the alert to the right runbook, with severity in one line each.
- Reading dashboards under pressure — four questions, habits that prevent wrong conclusions, what graph shapes usually mean.
- 10 scenario runbooks - 5xx spike, bad deploy (with a rollback safety check), database connections exhausted, slow queries, memory leak / OOM, disk full, queue backlog, dependency outage, certificate / domain / token expiry, missed cron job. Each: symptoms → check → mitigate → verify → escalate → prevent, read-only commands first.
- Escalation and update templates, 10-minute shift handover, alert hygiene from the on-call seat, after-the-incident checklist and mini-postmortem.
- On-call log (CSV) and a **worked example** of one night on call.
Who it's for: developers new to on-call, and teams who want a shared playbook without a dedicated SRE team.
Formats: Markdown, CSV. Example commands for Linux, Docker, Kubernetes, PostgreSQL (queries verified on PostgreSQL 16), Redis. **License:** Personal & Team. **Updates:** free for 12 months.