An organization has had a series of recent failures in one of their core applications.

TheThird Wayin DevOps emphasizescontinual learning and experimentation. One of the most effective resilience-building practices in this category ischaos engineeringor deliberate fault injection. This involves introducing controlled failures into systems—such as shutting down services, simulating network latency, or exhausting resources—totest the organization’s recovery capabilityin realistic scenarios.

By practicing recovery under controlled conditions, teams identify weaknesses in monitoring, automation, documentation, and communication before a real incident occurs. This improvesMean Time to Recovery (MTTR)and overall system resilience.

OptionA(constraint analysis) relates to theFirst Way(flow optimization). OptionCfocuses on feedback loops (Second Way). OptionD(adding redundancy) is a possible improvement but does not inherently practice and improve operational recovery skills in the same way as deliberate fault injection.

Therefore,Bis the most appropriate Third Way principle for improving resilience in this scenario.