Failure in the cloud is not a question of “if”; it is a question of “when.” From transient disk I/O latency to full-scale regional outages, failure in distributed systems is not only inevitable, but it is also often unpredictable. In response, enterprise architecture has matured from merely optimizing for uptime to actively engineering for resilience. This shift is not just technical; it is strategic. Availability and recoverability now shape board-level decisions, influence regulatory compliance, and determine business continuity in a globally connected economy.