Every enterprise has a brownfield morass of legacy software that isn't exciting anymore, but still runs in production. These workloads are often business-critical, but receive little investment outside of what it takes to keep the lights on. As SREs, these workloads can be particularly annoying - legacy platforms and code without instrumentation often mean our only source of observability data comes from logs. In this talk, I'll show you how you can transparently use cloud-native tricks to instrument an application without touching the application's code, whether it runs on a VM, a container, or elsewhere. Then, I'll show you how the data collected can be used strategically to generate the four golden signals of observability that we as SREs leverage to keep production online. If you're managing the morass of legacy software in production, this talk is for you.
Tyler Schade is a Distinguished Engineer at GEICO, where he leads software engineering and architecture for GEICO's infrastructure. An avid proponent of open source software, Tyler currently serves on the Istio steering committee and as a maintainer for a number of CNCF-hosted networking and security projects. Prior to GEICO, Tyler held roles at Solo.io, Ironnet, and Comcast, building deep networking, security and distributed systems experience. He lives in Philadelphia with his wife and their newborn.
