Cluster API v1.12.0 adds in-place updates and chained upgrades, so machines can swap parts without going down, and clusters can jump versions without drama.
KubeadmControlPlane and MachineDeployments now choose between full rollouts or surgical patching, depending on what changed. The goal: keep clusters stable, upgrades smooth.
Bigger picture: Cluster API is edging closer to what real workload orchestration should feel like, a smart balance between solid-state infra and lifecycle agility.
