Disaster Recovery

Automated Backup Strategy

Hypocrates database instances run continuous transaction write logging combined with automated daily snapshots. Backups are encrypted at rest with unique AWS KMS keys and are retained for a minimum of 30 days to facilitate granular point-in-time recovery.

Geographic Redundancy & Replication

To protect database instances against physical infrastructure failures, backup snapshots are replicate dynamically across separate geographic cloud zones (e.g., replicating database backups from US-East primary regions to US-West secondary environments).

Recovery Objectives (RPO & RTO)

Our disaster recovery framework is audited to achieve the following performance targets:

• Recovery Point Objective (RPO): Maximum 24 hours. In the event of a catastrophic regional failure, database states can be restored with a maximum data-loss interval of less than 24 hours.
• Recovery Time Objective (RTO): Maximum 4 hours. Critical systems, scheduler APIs, and patient records access are restored to full operational levels within 4 hours.

Restoration Drills & Validation

Our operations team conducts biannual disaster recovery dry-runs. These drills test the restoration of database snapshots to blank staging clusters to verify backup validity, process speeds, and engineering readiness.