Drifting into Fragility

This talk will look at complex systems failure analysis and how to apply it to system administration through the books Drift into Failure by Sidney Dekker and Antifragile by Nassim Taleb. There will be a focus on using postmortems to gain a better understanding of how complex systems fail and then coming up with strategies based on real-world examples to change the way that we run systems so that they are more understandable, less prone to failure, and easier to repair when outages do occur.

Matt Provost is the Systems Manager at Weta Digital. Weta Digital is a five-time Academy Award–winning visual effects facility in Wellington, New Zealand. The Systems team at Weta is responsible for all of the company's servers, storage, and networking. They run a 49,000 core renderwall. Matt has been a system and network administrator for over 15 years. He has a B.A. from Indiana University, Bloomington.