If there's one thing that's certain when you're building distributed applications, is that, sooner or later, one of those failure points scattered around your architecture will, well, fail. The only question that remains: is your application ready for it?

What happens to your precious data if Kubernetes kills the pod halfway through processing? How about a network hiccup? Will your API be able to handle client retries and retain consistency?

With this talk, I hope to get you to think about the resilience of your systems, go through a bunch of scenarios to remind you that failure is always an option, and that is fine, as long as you diligently design your applications.