Cluster API v1.12 Introduces In-place Updates and Chained Upgrades
The Kubernetes community has released Cluster API version 1.12.0, a significant update designed to streamline the declarative management of Kubernetes cluster lifecycles. This release introduces two major features: in-place updates and chained upgrades, aimed at reducing friction in common operational tasks. Traditionally, Cluster API relied on immutable infrastructure principles, where machine changes triggered rollouts by creating new machines and deleting old ones. While predictable, this approach could cause workload disruption. The new version allows the system to automatically choose between immutable rollouts and in-place updates based on the specific change required. By leveraging new update extensions, users can now modify existing machines without recreation, minimizing downtime and resource churn. Both KubeadmControlPlane and MachineDeployments support these capabilities. The update emphasizes simplicity and usability, ensuring that platform teams can define desired states while controllers reconcile them with minimal manual intervention. This evolution represents a strategic shift towards balancing the reliability of immutable infrastructure with the efficiency of in-place modifications, significantly expanding the operational boundaries for managing control plane machines and worker nodes in diverse environments.
Wire timeline
Cluster API v1.12 Introduces In-place Updates and Chained Upgrades
The Kubernetes community has released Cluster API version 1.12.0, a significant update designed to streamline the declarative management of Kubernetes cluster lifecycles. This release introduces two major features: in-place updates and chained upgrades, aimed at reducing friction in common operational tasks. Traditionally, Cluster API relied on immutable infrastructure principles, where machine changes triggered rollouts by creating new machines and deleting old ones. While predictable, this approach could cause workload disruption. The new version allows the system to automatically choose between immutable rollouts and in-place updates based on the specific change required. By leveraging new update extensions, users can now modify existing machines without recreation, minimizing downtime and resource churn. Both KubeadmControlPlane and MachineDeployments support these capabilities. The update emphasizes simplicity and usability, ensuring that platform teams can define desired states while controllers reconcile them with minimal manual intervention. This evolution represents a strategic shift towards balancing the reliability of immutable infrastructure with the efficiency of in-place modifications, significantly expanding the operational boundaries for managing control plane machines and worker nodes in diverse environments.
Kubernetes Blog