Kubernetes can support clusters with thousands of nodes, but just because you can scale that far doesn't mean you should. As clusters grow, so do operational complexities—ranging from slower upgrades and API server strain to expanded blast radius and multi-tenancy challenges.

This talk provides a framework for thinking strategically about Kubernetes cluster scale. It examines key trade-offs, including control plane performance, failure domains, and upgrade complexity, while addressing the challenges of securely supporting multiple tenants and evaluating when multi-cluster strategies make sense. Rather than focusing on theoretical limits, the session emphasizes a pragmatic approach to scalability, aiming to help teams design clusters that are not only large, but resilient, efficient, and maintainable.