Serverless technologies offer a great foundation for building resilient applications that can withstand much turbulence in the production environment. For instance, AWS Lambda automatically deploys your code to three availability zones and replaces faulty virtual machines on the fly. Despite this, there are still many other types of failures that can still affect our application. Perhaps there is an outage with a third-party service we depend on, or maybe a sudden surge in throughput has pushed us over the throughput limit and caused some user requests to be throttled.

What can we do to minimize the impact of these failure modes on the user experience and prevent data loss?

In this talk, Yan will look at some common patterns (and trade-offs) for building resilient serverless applications.