Tales from the Trenches: How Caches Fail at Scale

Caching is supposed to make everything faster — until it doesn’t. Madelyn, a Principal Engineer at Amazon ElastiCache, has spent years keeping high-performance caching systems alive under the most demanding production conditions on the planet. In this talk, she’ll share what actually goes wrong at scale: from request cascades that turn a minor traffic spike into a full-blown outage to large-scale failure modes that no architecture review ever anticipated. In this talk you’ll learn how to thing about failure modes for high performance systems so that you can anticipate and react quickly when things go wrong.