Offline RKE1/RKE2 node reported as "Running" in Cluster Management and as "In Progress" on Kube Nodes page

[thehejik]

For both RKE1 and RKE2 clusters I tried to shutdown one of the nodes (on cluster with 3 all-roles nodes) and in Cluster Management the offline node is still reported as Running after 10+ minutes.

Under EXPLORE CLUSTER/Cluster->Kube Nodes the offline node is visible as In Progress which is a bit better but still doesn't correspond with reality.

Seen on rancher/rancher:master-head cf5a75f

[sowmyav27]

Similar bug - rancher/rancher#32891

[gaktive]

Working as designed since Kube Nodes will have a different state than "regular" nodes identified in the first screenshot. However, if this proves confusing based on usage, we may need another indicator.

Ideally, that 10 minute period should not happen but it should sort itself out.

[thehejik]

On Kube Nodes page the node are at least marked with status In Progress what is acceptable from my point of view.

Problem is that the node state is Active even 8 hours after a node shutdown in Cluster Management/Node Pools view and won't switch to something like Offline. I consider this as a bug we should fix.