Node Manager - How Yahoo Manages Thousands of Nodes at Scale

Learn how Yahoo manages thousands of Kubernetes nodes at scale through their custom Node Manager operator in this 41-minute conference talk from the Cloud Native Computing Foundation. Discover Yahoo's approach to operating 37 on-premises Kubernetes clusters hosting 2,700 applications across more than 8,000 physical nodes, and explore the significant challenges they face in daily cluster management including Kubernetes version upgrades, security updates, OS version upgrades, and handling node failures. Understand how Yahoo developed the Node Manager operator to automate maintenance tasks, perform controlled Kubernetes version upgrades for both control plane and worker nodes without impacting running applications, and conduct continuous health checks with automatic node remediation based on nodegroup configurations. Gain insights into large-scale infrastructure automation strategies that reduce manual engineering work and improve operational efficiency in enterprise Kubernetes environments.