The On-Call Runbook Every Small DevOps Team Needs (Free Template)

When the pager goes off at 3am, you need a runbook that tells you exactly what to do. Not a 200-page document nobody reads. A one-page runbook that gets you from alert to resolved as fast as possible.

This template is designed for teams of 1-10 people who run production without a dedicated SRE team.

When You Get Paged

1. Acknowledge within 5 minutes - tells the team you're on it
2. Check alert details - what service, what metric, what threshold
3. Check the dashboard - is this real or a fluke? Look at trends
4. Check recent changes - deployments in the last 2 hours?
5. Determine severity - SEV1 (business down), SEV2 (degraded), SEV3 (minor)
6. Create incident channel if SEV1 or SEV2

Quick Fixes by Symptom

Service Down (502/503)

1. Check if process is running
2. Check recent deployments
3. Roll back if recent deploy caused it
4. Check resource limits
5. Check logs
6. Restart if needed

High Error Rate (5xx)

1. Check error logs for patterns
2. Check database connections
3. Check external API status pages
4. Check for config changes
5. Roll back if bad deploy

High Latency

1. Check CPU/memory usage
2. Check database slow queries
3. Check network/DNS
4. Check queue depth
5. Scale up if needed

Database Issues

1. Check connection pool
2. Check slow queries
3. Check disk space
4. Check replication lag
5. Kill long-running queries if blocking

Disk Space Full

1. Check disk usage
2. Find large files
3. Check log files
4. Rotate/compress old logs
5. Clean up temp files

Escalation Paths

Severity	Who	Response Time
SEV1	On-call + Lead + CTO	5 min
SEV2	On-call + Lead	15 min
SEV3	On-call only	Next day

Communication Template

[SEV2] [Service] experiencing elevated error rates
Started: [timestamp]
Impact: [who/what is affected]
Status: Investigating
Next update: [15 min from now]

Update every 15 minutes. Even if the update is "still investigating."

Post-Incident

1. Schedule post-mortem within 48 hours
2. Write timeline of events
3. Identify root cause and contributing factors
4. Create action items with owners and deadlines
5. Update this runbook with learnings

Pro Tip

Keep this runbook next to your pager. When you're half-awake at 3am, you don't want to search for the runbook. Print it. Pin it. Use it.

---

Get the complete runbook with 20+ scenarios: Ops Starter Kit

Free resource: The First 30 Minutes - free incident quick-start checklist.

Use code LAUNCH50 for 50% off anything in the store.