Automating Kubernetes Cluster Updates: Achieving Zero Downtime Effortlessly
Talkat KubeCon + CloudNativeCon Europe 2025 · on Thursday, 2025-04-03· at 17:30· Level 0 | ICC Capital Hall | Room I·
summary
Automates Kubernetes control-plane and node upgrades using Karpenter to replace nodes with zero downtime on AWS, Azure and AlibabaCloud, with examples of rolling-replace vs node-pool replacement and a demoed limit of simultaneous replacements controlled by disruption budgets. Describes limits: no gradual intermediate-state control and no built-in rollback, and extends Karpenter via CloudPilot AI for intelligent node selection and ML-driven spot interruption prediction up to 120 minutes with automated migration strategies.
AI-written summary · about
chapters (15)
- 0:00Kubernetes updates introduction
- 1:04Why data plane updates matter
- 2:10Risks of manual updates
- 3:40Master and node upgrade process
- 4:47Rolling update strategy
- 7:19Kapanda for automated updates
- 8:46Simplifying node upgrades
- 11:01Node disruption budget
- 12:53Kapanda limitations today
- 13:36Kapanda cross-cloud support
- 14:40Current Kapanda drawbacks
- 15:45Extending Kapanda with Cloud Palai
- 16:08Intelligent node selection
- 16:29Spot interruption prediction
- 17:13Summary and resources
related talks
19mKarpenter, the Future of Kubernetes Node Management
22mUse Karpenter to scale, optimize & upgrade K8s
22mHarnessing Karpenter: Transforming Kubernetes Clusters with Argo Workflows
22mKarpenter Efficient scaling of Kubernetes clusters
33mNavigating and Optimizing Karpenter - Avoiding Common Pitfalls
01h 01mJust-in-time Nodes for Any AWS EKS Cluster: Karpenter
27mHarness the power of Karpenter to scale, optimize & upgrade K8s
25mAutomating Kubernetes Upgrades for Zero Downtime