“Uber Scale. Stories and lessons from the history of scaling Uber SOA”
Christopher Adams. Uber
In this talk, Chris will discuss how Uber scaled several key components of its SOA infrastructure, and extrapolate some lessons and useful strategies that other companies can apply as they scale.
- Compute platform – how Uber’s infrastructure team brought its customers along on the journey to the platform they truly needed.
- Dispatch system – how Uber’s dispatch engineering team proved Joel Spolsky wrong by rewriting Uber’s most important system from scratch – without anyone noticing.
- Developer productivity – how Uber’s developer productivity team decided when to invest in its own tools and frameworks, when to start from off the shelf tooling and customize, and the tipping points that made these decisions right.
Originally from Melbourne, Australia, Chris has spent over 10 years as a software engineer in a variety of roles up and down the stack. Graduating from Computer Science in Melbourne in 2005, he used his spare time to co-found an organization which ran Australia’s largest bring-your-own-computer networked PC gaming events. But Chris always dreamed of moving to Silicon Valley. In 2013 when he moved, he discovered a much wider world: at Pivotal Labs he discovered agile and pair programming. Soon after he moved into a role in infrastructure engineering and discovered his passion. His next move was to Uber, where he continued his infrastructure engineering passion through his involvement in a rebuild from scratch of Uber’s dispatch system as well as the creation of Uber’s SRE (Site Reliability Engineering) team. Today, Chris continues to lead the push for reliability thinking within Uber’s engineering team of thousands. Chris brings with him to ICSOC 3 years of experience of Uber hypergrowth, and he has seen Uber change from a chaotic startup into an organization where great numbers of people work together at scale to achieve shared goals.