KubeCon + CloudNativeCon North America 2024

Cash App's Compute team is responsible for the health and maintenance of the company's Kubernetes clusters, and the enablement of service owners to deploy their services into these clusters with confidence. Over the past year, we've made strides in improving our reliability and uptime, part of which involved introducing a paradigm around creating new Kubernetes clusters in our service ecosystem that allow us to seamlessly transition services in/out of to simplify cluster upgrades and provide us with guardrails against common outages. This talk intends to walk you through our experience introducing new Kubernetes clusters for our services at Cash App, migrating and splitting service traffic across clusters with zero downtime, and thinking through tooling adoption / creation to simplify cluster maintenance as our overhead scales.