- Published
- Updated
- Sources checked
A successful backup notification shows that a scheduled job completed. It does not establish that study records can be restored with their context, permissions and audit history after an interruption.
Define recovery priorities
Define recovery priorities for each system with business owners. Identify acceptable data loss and recovery time as study-specific operational objectives. Consider dependencies such as encryption keys, identity services and vendor export tools; a data file alone may be unusable.
Restore an isolated study copy
Restore a representative study dataset into an isolated, authorised environment. Compare record counts, attachments, metadata and traceability with the agreed test source. Check whether users can retrieve the restored content without accidentally sending live notifications or changing production records.
Restoration checks
- Record backup scope, encryption and custody responsibilities.
- Test restoration with documented acceptance criteria.
- Assign corrective actions for missing dependencies or failed comparisons.
Accept reproducible recovery
Accept the exercise when the restored evidence meets the defined objectives and the recovery steps are reproducible. Retain the test version, results and deviations. Repeat testing after meaningful architecture changes and use findings to update the recovery plan. Check that unauthorised users cannot read the restored copy, and assign custody or controlled removal of temporary test data after the exercise.
Frequently asked questions
- Is backup the same as archiving?
- No. Recovery copies and long-term retained records serve different purposes and need distinct controls.
- Must every system use the same recovery target?
- Set targets from study impact, dependencies and feasible recovery arrangements.