Stopping Bugs from going to production at Uber

The Uber office in Aarhus houses 80+ engineers, many of which are building planet-scale infrastructure. I work on the deployment platform called Up which runs almost 6K micro-services across more than 12 datacenters in 3 geographical regions and doing about 10K deployments weekly. DevOps is all about going from development to production faster / more often, with less effort, and more confidence – because the steps involved have been automated. Part of the confidence comes from avoiding manual steps, where mistakes can be made. But another part is the safety that can be achieved by catching code- or business-level errors as early as possible in the process. In this talk, I will focus on what is done in Uber to prevent bugs from going all the way from development to production and potentially cause an outage. I will share a glimpse into the DevOps experience in Uber and how we have achieved a high level of automation and safety.



Ulrik Skyt

I am a senior software engineer in the Aarhus Uber office, where I have worked for almost 3 years as part of a strong team that works on Uber’s next-generation service management platform, called Up.