The correct answer is C, 'Design for automated recovery from failure.' This principle is part of the AWS Well-Architected Framework's Reliability Pillar, which emphasizes the importance of building applications that can automatically recover from failures without causing significant disruptions. This is in contrast to options A, B, and D, which do not align with the Well-Architected Framework's best practices. Option A suggests keeping static data closer to compute resources, which is not a principle of the framework. Option B, provisioning for peak capacity, can lead to over-provisioning and inefficiency. Option D, using tightly coupled components, can create dependencies that make the system less resilient and harder to manage. The Well-Architected Framework encourages the opposite approach: designing for failure, using automation to manage resources efficiently, and keeping components loosely coupled for better scalability and maintainability.