Observability for Emerging Infra: What Got You Here Won't Get You There

Distributed systems, microservices, containers and schedulers, polyglot persistence... modern infrastructure is ever more fluid and dynamic, chaotic and transient. Likewise, individual engineering roles can no longer be broken down neatly into software engineers (who write the code) and ops engineers (who deploy the code (and buffer the consequences)). Many teams have already sailed past an event horizon of complexity and found that their old tools and processes no longer work for them. But why, exactly? What was wrong with traditional metrics and logs? Why are they failing to keep pace with modern requirements? Isn't observability just a marketing term for monitoring? And what on earth can we do about it? In this talk, we'll cover the shortcomings of traditional metrics and logs, and the technical and cultural differences between monitoring and observability. We'll also talk about the deep cultural revolution underway from siloed specialties towards software ownership (in every type of engineering role)—and what exactly does that mean when it comes to systems observability?—as well as the technical practices and mental shifts that are absolutely required to keep pace with modern infrastructure.