Training material
Alertmanager and karma hands-on
Alerts, silences and dashboards
From the first alert to the control room. You write Prometheus rules for notes-api, build a routing tree with teams and on-call, shape mails with templates, suppress follow-up alerts with inhibition, silence maintenance windows and run three Alertmanagers in an HA cluster. Then karma brings two environments onto one screen, with filters, multi-grid, silences and acknowledgements, login via Caddy, silence ACLs and healthchecks on the watchdog. Finally everything runs on k3s with kube-prometheus-stack. Every failure in the book really happened during testing.
- Alertmanager
- karma
- Prometheus
- Observability
- Kubernetes
Training plan
- Foreword
- Day 0: The Training Ground
- Day 1: Alerts Are Born
- Day 2: Alertmanager
- Day 3: The Routing Tree
- Day 4: Receivers and Templates
- Day 5: Creating Quiet
- Day 6: High Availability
- Day 7: karma
- Day 8: Filters and View
- Day 9: Multiple Alertmanagers
- Day 10: Silences in karma
- Day 11: Access
- Day 12: History, Health Checks and Troubleshooting
- Day 13: karma on Kubernetes
- Day 14: Operations
- Day 15: Competition
- Appendix A: Cheat Sheet
- Appendix B: Final Exam