Theoretical Foundations of Software Resilience Engineering for Distributed Systems in Sub-Saharan Africa

Distributed software systems deployed in Sub-Saharan Africa face resilience challenges qualitatively different from those assumed in dominant fault tolerance and resilience engineering frameworks, necessitating original theoretical work that situates resilience engineering in African infrastructure realities. This dissertation develops a theoretical foundation for software resilience engineering specific to distributed systems in Sub-Saharan Africa, drawing primary empirical grounding from Nigerian distributed systems deployments across three geopolitical zones. A multi-paradigm research methodology combines a systematic literature review (80 papers), grounded theory fieldwork at 10 Nigerian organisations operating distributed software systems in Enugu (South East), Kaduna (North West), and Benin City (South South), and design experiments implementing and evaluating three candidate resilience architectural patterns. The fieldwork reveals that the dominant resilience challenges in Nigerian distributed systems are infrastructure-induced: power outages causing cascading node failures, MTN and Airtel network partitions causing unpredictable split-brain scenarios, and third-party API unreliability at rates 4 to 8 times higher than international benchmarks. These challenges are inadequately theorised in existing resilience frameworks such as CAP theorem applications and chaos engineering methods developed for cloud-native Western infrastructure. The dissertation introduces the Infrastructure-Hostile Distributed Systems (IHDS) theory, positing that African distributed systems engineering requires a distinct resilience paradigm centred on graceful degradation hierarchies, asynchronous-first communication design, and infrastructure hostility as a first-class design parameter. Keywords: software resilience, distributed systems, Sub-Saharan Africa, infrastructure, fault tolerance