Scaling to Infinity: The Quest for Fully Automated, Scalable, Self-Healing Systems

The number of servers in our systems is constantly increasing. With the move to microservices and continuous deployment practices, deployments frequency is higher than ever, yet we are expected to handle increased demand with the same workforce as before. How can we operate clusters at scale with minimum human involvement? The answer lies in adoption of new practices and tools. Join Viktor Farcic as he discussions how container schedulers are becoming a de facto standard for deploying services at scale and continuous deployment is the new norm for complete automation of software delivery. You'll learn about the problems encountered while building systems that operate at scale and understand how to create a cluster that can be scaled to (almost) any size. Viktor will discuss how to enable self-healing and self-adaptation to build a reliable, fault-tolerant system. You'll discover the metrics necessary to control any type of automated cluster management and leave with tips and techniques for how to successfully combine these concepts into a self-sufficient system that can operate at any scale.