Data Localisation
& Cloud Repatriation
Modern data localisation is a system-boundary problem—not merely a database-location problem.
Programme context
Trace the whole system.
Nigeria’s evolving financial-sector data-localisation requirements create engineering questions for banks, fintechs, payment providers and other regulated institutions.
Data may appear in databases, queues, logs, traces, caches, backups, analytics systems and support tooling. Evaluation begins with data flows, workload dependencies and failure boundaries.
The Lab does not label an architecture “CBN compliant.” Architecture alone does not establish regulatory compliance.
Questions under test
Where does regulated data actually flow?
What happens when application compute and databases are geographically separated?
How much latency does split-cloud architecture introduce?
How do chatty applications behave across international links?
What happens when cross-border connectivity fails?
Which workloads need to move closer to the data?
How should backup and disaster recovery be designed?
How do logging, analytics and observability affect residency boundaries?
What operational complexity does hybrid infrastructure introduce?
What are the cost implications of cross-cloud traffic?
Reference workload · planned
Ingress
API Gateway
Synthetic requests
Transaction plane
Account + Payment
Services + PostgreSQL
Asynchronous plane
Fraud + Notification
Event consumers
This reusable workload is planned and has not yet been implemented. It will use synthetic data only—never customer or production financial data.
Programme experiments
Conceptual topology
Baseline pending
Data localisation
Split-Cloud Transaction Architecture
Measure the consequences of running application compute outside Nigeria while maintaining transaction data on Nigerian infrastructure.
Conceptual topology
Baseline pending
Resilience
Cross-Border Failure Simulation
Understand how a split-cloud financial workload behaves when connectivity between foreign compute and Nigerian infrastructure is degraded or unavailable.
Conceptual topology
Baseline pending
Data localisation
Workload Localisation Patterns
Compare placement strategies for transaction-critical services, persistent data and supporting workloads.