How does failure-recovery (software bugs, hardware, or otherwise) or resilience-engineering more generally work on a system that demurs isolation and redundancy

How does failure-recovery (software bugs, hardware, or otherwise) or resilience-engineering more generally work on a system that demurs isolation and redundancy (honest question)?