Currently there seems to be no documentation on recommended approaches to node OS patching and whether replacement is the recommended approach over patching in-place.
For example
buid/replace with rke
cordon and patch in-place
health checks for things like etcd
Node types
Rancher local cluster
Downstream controlplane/etcd/worker build/destroy where possible
Currently there seems to be no documentation on recommended approaches to node OS patching and whether replacement is the recommended approach over patching in-place.
For example
Node types