Backup is a technical control; recovery is an operational process. Green checkmarks in a console are not enough when systems must be restored under pressure.
Separate production data and backups
Backups reachable with the same administrative paths as production may be affected by the same incident.
- multiple backup copies
- offsite or separated target
- restricted backup administration
- consider immutable protection
Translate RPO and RTO into clear objectives
Define acceptable data loss and recovery time per service rather than for “IT” as a whole.
- prioritize critical systems
- define acceptable data loss
- define recovery time
- consider dependencies
Restore tests belong to normal operations
Tests verify that data is readable, credentials are available and recovery instructions work.
- file restore
- VM/system restore
- application start test
- document results
Recovery must not depend on one person
Documentation and technical cover reduce organizational single points of failure.
- document access
- write recovery steps
- define technical cover
- set escalation contacts