View a markdown version of this page

LSREL09-BP03 Verify data integrity and point-in-time recovery - Life Sciences Lens

LSREL09-BP03 Verify data integrity and point-in-time recovery

Protect application and data state so that you can restore to a precise point-in-time prior to a failed change. Backups must be consistent, validated, and aligned with organizational RTO and RPO targets. Include configuration and metadata in backups so restored environments are audit-complete.

Desired outcome: Data and application state can be restored to the exact state prior to the failed change, preserving adherence and operational continuity.

Common anti-patterns:

  • Infrequent or unplanned backups without restore validation.

  • Omitting metadata or configuration in backup sets.

  • Not aligning backup frequency to defined RTO and RPO needs of regulated systems.

Benefits of establishing this best practice:

  • Avoids loss of experiment data, patient records, or audit evidence during failed changes.

  • Shortens recovery time, preserving study timelines and sample integrity.

  • Demonstrates recoverability to auditors and stakeholders.

Level of risk exposed if this best practice is not established: High

Implementation guidance

Define RTO and RPO aligned to the criticality of datasets and regulated processes. Implement automated, policy-driven backups that include application-level snapshots, database backups (including transaction logs for point-in-time recovery), and configuration exports. Regularly perform restore drills and document restoration time and fidelity as part of validation evidence. Store backups in tamper-evident, redundant storage and that retention policies meet regulatory retention requirements.

Implementation steps

  1. Implement policy-driven backups using AWS Backup for supported services.

  2. Enable automated backups and snapshots for databases (for example, automated backups for Amazon RDS and snapshot schedules for Amazon EC2 or Amazon EBS).

  3. Configure point-in-time recovery for services that support it, such as enabling PITR for Amazon DynamoDB.

  4. Store and catalog backup artifacts in Amazon S3 with appropriate lifecycle and retention rules, and use AWS Backup's recovery testing features to validate restores.

Resources

Related best practices:

  • Long-term storage and reliable recovery of trial data

  • Observability and monitoring of pipelines

  • Automated validation in deployments