DIREKT Backup and Recovery

Objectives

Final RPO/RTO set before pilot. Initial planning targets:

  • core database RPO ≤ 24 hours, with managed point-in-time capability preferred;
  • pilot restore target ≤ 8 hours;
  • production targets tightened from measured business impact.

Coverage

  • PostgreSQL;
  • object storage/evidence;
  • configuration/versioned rules;
  • infrastructure definitions;
  • signing/secret recovery procedures;
  • Pages/repository already versioned.

Controls

  • automated backups;
  • encryption;
  • separate failure domain/account where feasible;
  • retention;
  • access logging;
  • restoration documentation;
  • integrity checks;
  • no untested “backup exists” claim.

Restore exercise

At least before pilot and periodically:

  1. create isolated restore environment;
  2. restore database and storage references;
  3. run integrity/application checks;
  4. verify sample evidence access authorization;
  5. record time/gaps;
  6. destroy restored sensitive data securely.

Disaster

Use docs/operations/DISASTER_RECOVERY.md for decision and communication flow.