On-call handbook
> The handbook is what a new team member reads before their first shift — and what an experienced operator skims when context-switching at 2am.
On-call handbook
The handbook is what a new team member reads before their first shift — and what an experienced operator skims when context-switching at 2am.
Who we are
| Team | |
| Services supported | |
| On-call schedule | |
| Primary contact method | |
| Escalation path |
Before your shift
- Confirm you have access to: dashboards, logs, alerts, incident channels
- Read any new runbooks since your last shift
- Check for ongoing incidents or maintenance windows
- Know who to escalate to for each service tier
During your shift
Alert response SLA
| Severity | Response time | Update frequency |
|---|---|---|
| SEV-1 | 15 min | Every 30 min |
| SEV-2 | 30 min | Every 60 min |
| SEV-3 | 4 hours | Daily |
| SEV-4 | Next business day | N/A |
Acknowledging an alert
- Ack within the SLA window.
- Check the runbook for the symptom (see
docs/internal/operations/runbooks/). - If no runbook exists or it’s stale, triage with standard tools and file a follow-up to write it.
- Post in the incident channel: service, symptom, severity, estimated impact.
If you need help
| For this | Contact |
|---|---|
| Runbook exhausted | Service owner |
| Security suspicion | Security team (#security) |
| Data loss risk | DBA / platform team |
Standard tools
| Tool | URL / access | What it’s for |
|---|---|---|
Handover
- Summarise ongoing incidents in the handover doc
- Flag incomplete investigations
- Confirm next on-call is awake and acknowledged the handover
After your shift
- File runbook improvements for anything that was missing or wrong
- Update the handbook if any tool or contact changed
- Log any “first time” issues you hit — they are the next rota’s risk
Related
- Runbook index:
- Incident retrospective template:
- Escalation matrix:
- Dashboard collection: