Architect for failure and you’ll build resilience into your system. When AWS’s US-East-1 region went dark in late October, followed just a week later by a Microsoft Azure outage, it was yet another stark reminder that even the world’s biggest cloud vendors are not immune to failures. A simple DNS failure in AWS’s Route 53 rippled outward, knocking out applications, disrupting database services, and reminding us how dependent our tech infrastructure has become on a handful of cloud regions. With “an inadvertent tenant configuration change,” the Azure outage further highlighted the instability of some of these systems, once again demonstrating how small changes can have quite a large impact.