Surviving From Endless Issues in 7K+ Kubernetes Clusters
Explore how a small DevOps team at Kakao Corp manages to handle issues arising from over 7,000 Kubernetes clusters in this insightful conference talk. Learn about the challenges faced by the private Kubernetes as a Service team as their infrastructure grew from 2,000 clusters with 20,000 nodes to over 7,000 clusters with more than 100,000 nodes in just one year. Discover the strategies, tools, and techniques employed to effectively manage the increasing number of on-call issues. Gain valuable insights into the team's approach to problem-solving, including the development of custom tools and the decision to open-source some of their solutions. This presentation offers practical knowledge for DevOps professionals dealing with large-scale Kubernetes deployments and those interested in efficient issue management in rapidly growing infrastructures.