Keeping the service and the team running through supplier, infrastructure and people disruptions.
01Approach
- The full platform is described as code, so it can be rebuilt in another AWS availability zone, or another region, from the repositories and backups.
- Critical knowledge is written down as runbooks next to the code, not held by individuals.
- Deploy pipelines can run on an alternative build provider without code changes.
- Key suppliers are chosen for their own redundancy (multi-zone AWS, Cloudflare’s global network).
02Scenarios
| Scenario | Response |
|---|---|
| Loss of a database node | Automatic failover inside the replica set |
| Loss of a server | Rebuild from template, reattach data volume or restore snapshot |
| Faulty release | Switch back to the previous version kept on the server |
| Build provider disruption | Switch to the alternative build provider |
| Key person unavailable | Runbooks and least-privilege roles let another operator act |
03Contact
Questions about this policy, security questionnaires and Data Processing Agreement requests: privacy@mentornity.com.