LSREL08-BP03 Align architecture priorities with scientific and regulatory context
Early in the design cycle, evaluate the scientific and regulatory priorities of each workload. In patient-facing clinical workloads, design for continuous availability with reconciliation workflows. In regulated manufacturing, prioritize data integrity in the architecture, even if this limits availability. Planning with these priorities in mind avoids trade-off surprises later.
Desired outcome: Availability trade-offs are consciously designed to meet workload-specific requirements.
Common anti-patterns:
-
Treating each workload with a uniform HA design.
-
Prioritizing availability at the cost of data integrity in regulated processes.
-
Failing to document the rationale for architectural trade-offs.
Benefits of establishing this best practice: Aligns architectures with the unique regulatory and scientific priorities of each workload, and improves transparency in audits.
Implementation guidance
Architectural planning should explicitly evaluate the workload's scientific and regulatory context before finalizing availability and failover mechanisms. For example, clinical trial systems may tolerate some reconciliation steps post-recovery if availability is critical, while batch manufacturing systems may require uncompromising data integrity even if it reduces availability. Document these trade-offs, rationale, and validation approach as part of the design package.
Implementation steps
-
Capture availability and data integrity requirements in system requirements specifications and risk assessments.
-
Map RTO and RPO targets to these requirements.
-
On AWS, implement availability features with services like Amazon Aurora Multi-AZ or Amazon S3 Versioning, paired with validation steps for data integrity.
-
Store architecture trade-off documentation and validation outcomes in a controlled repository for audit readiness.
