Explore best practices for setting up etcd with Kubernetes to host clusters with thousands of nodes in this informative conference talk. Learn from tech leads at Datadog and Google Cloud as they share their expertise on etcd deployment architecture and configuration for large-scale clusters. Discover strategies for optimizing performance, reliability, and scalability in etcd instances, particularly for AI/ML/HPC workloads and internet-scale serving applications. Gain insights into handling disk I/O and network throughput bottlenecks, managing API server restarts, and mitigating their impact on etcd. Benefit from the practical experience of industry leaders who have successfully implemented and managed Kubernetes clusters with thousands of nodes for years.