Operations
By Rafael Duarte · created 10 Jan 2022, 08:45 UTC · last updated 21 Oct 2026, 15:05 UTC · version 9
space-homeoperations
How we keep production running and recover when it doesn't: backups and restores, monitoring, the status page, planned changes and what we learn from incidents. Engineering's runbooks for specific failures are in the Engineering space.
Owner: Oskar Lund (site reliability) since June 2026. Before that, Rafael Duarte and Mira Holm.
Pages
| Page | What |
|---|---|
| Operations runbook | backups, restore drills, disaster recovery, SAML certificates |
| Monitoring and alerts | what we watch and what pages someone |
| Status page | what customers see and who posts |
| Change calendar | planned production changes and freezes |
| Incident review process | how we review an incident |
| Incident summaries | what happened, for SEV1 and SEV2 |
Jira and Emargen
- Incidents, production changes and operational tasks are Jira issues in OPS. Incident, Task or Change.
- Incident reviews are written in Emargen from the Incident review template, and listed on Oskar's Incidents list.
- The Incident watch agent task runs every day at 08:30 and writes up what changed in the OPS project since the day before.
Being on call
The rota, severities and what to do when paged: On-call.
Change history
| Version | Date | Who | What |
|---|---|---|---|
| 6 | 22 Jun 2026 | Oskar | Oskar takes over; OPS project, incident reviews |
| 7 | 13 Jul 2026 | Oskar | Change calendar |
| 8 | 16 Sep 2026 | Oskar | Incident watch agent task |
| 9 | 21 Oct 2026 | Oskar | Kelderby's weekends on the change calendar |