GitOps Continuous Delivery at Scale with Flux - Stefan Prodan

In this session, Stefan will talk about best practices when running Flux at scale in production. He will start with an overview of the scaling capabilities of Flux controllers, from vertical scaling to horizontal scaling and sharding. He will do a deep dive into each scaling method and will explain when and how to use them in relation to multi-tenancy, cluster fleet size and workload complexity. Stefan will introduce the Mean Time To Production benchmarking tool that the Flux team has developed using CUE lang and Timoni. The benchmark measures the time it takes for Flux to deploy thousands of Helm charts and Kustomize overlays on Kubernetes clusters. He will explain the benchmark results and will share the lessons learned from running the benchmark on different Kubernetes distributions and providers. Stefan will conclude the session with a quick glance at the Flux roadmap and will share the plans for the next GA release.