Workload Localisation Patterns
Compare placement strategies for transaction-critical services, persistent data and supporting workloads.
- AUTHOR
- Hexcore Engineering Lab
- UPDATED
- 2026-10-04
- EVIDENCE STATE
- HYPOTHESISED / PLANNED
01 / Research question
Which workload placement patterns best balance latency, resilience, operational complexity and residency boundaries?
Hypothesis — not a finding
Moving transaction-critical services closer to their data will reduce network sensitivity while increasing local operational requirements.
02 / Architecture
Pattern A
Foreign application
Nigerian database
Pattern B
Local transactions
Foreign supporting workloads
Pattern C
Localised stack
Application + database
03 / Environment
- Pattern A
- Foreign application + Nigerian database
- Pattern B
- Nigerian transaction services and database + foreign support
- Pattern C
- More extensively localised application architecture
- Data
- Synthetic only; no customer or production data
04 / Methodology
A future reusable synthetic fintech workload. It is a planned test fixture and has not yet been implemented.
- 01Define identical functional boundaries
- 02Deploy each placement pattern
- 03Run the same workload profile
- 04Measure healthy-state behaviour
- 05Introduce equivalent failure scenarios
- 06Compare evidence without compliance claims
05 / Measurements
End-to-end latency
Throughput
Cross-boundary traffic
Failure isolation
Recovery characteristics
Operational complexity
Infrastructure cost
06 / Results & observations
This experiment has not yet been executed.
This space will contain measured results, observations and failure modes after repeated runs and methodological review. No benchmark values are available yet.
07 / Limitations
- —Reference workload is not yet implemented
- —Results will be architecture- and provider-specific
- —Regulatory interpretation is outside the experiment scope
08 / Reproduction & artefacts
Commands, configuration, dataset generators and raw data will be linked here when they exist.
Related 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.