Presentation: Tweet"S*%$ Happens: How You Build to Survive"
In a sufficiently complicated system things will go wrong, in a sufficiently large complex system things will go wrong all the time. Cloud computing is built upon some of the most well-designed, complex systems and has been built to anticipate these failures. Yet still there are outages. How do you build applications and processes that run happily in a SNAFU environment?
As developers there are certain steps/decisions we can make that buffer our apps against outages caused by new and "interesting" failure modes. The key is to not make decisions that block us from enjoying the scale and multi-region features of cloud compute while at the same time anticipating and embracing failure in application design so we can have fewer sleepless nights.
Download slides