Recently there was a blog post on Linkedin talking about the skills required to be a successful network engineer in the future. While the skills listed were all relevant and made sense, there was a key element missing – an operational mindset. To be successful at building and managing large scale infrastructure there has to be an operations first mindset to developing technology. It’s not simply about “automating” things or having a dev-ops team. This talk uses Facebook’s disaster recovery efforts as a case study to describe this mindset and approaches to deploying infrastructure that survives disasters without significant human intervention.