Major disruptions in cloud services have highlighted the vulnerability of businesses relying on online platforms. Fastly experienced a significant day on November 18th, marked by a widespread outage while simultaneously facing advanced DDoS attacks. Remarkably, during this chaos, several key customers shifted massive amounts of traffic—tens of petabytes—over to Fastly due to issues with another provider. Fastly managed to handle both the increased traffic and the attack without any disruption for end users. In response to rising outages, adopting a multi-vendor strategy has become essential for organizations. This approach leads to reduced downtime and enhanced resilience. By diversifying their infrastructure—using multiple ISPs, cloud providers, and CDNs—companies can better navigate incidents and ensure consistent user experiences. Fastly emphasizes that platforms must be designed with the understanding that failures will happen. Their systems are structured to maintain performance during unexpected demand spikes and to handle attack traffic while keeping legitimate users online. Fastly outlines several critical design principles for resilience: prioritizing availability and integrity, ensuring predictable performance, and adapting quickly to changes in demand. These capabilities stem from thoughtful architecture and ongoing improvements, rather than being reactionary measures. For companies reassessing their infrastructure strategy, Fastly offers guidance on building robust systems that prioritize high availability.