Inconsistent states between home dashboard, cluster management, and cluster explorer.

[phillipsj]

We have noticed that the UI doesn't report the same states across the different UIs on 2.6-head.

[vincent99]

This provides no evidence to suspect the UI is displaying anything other than what the API state says it is. You're looking at multiple separate resources which each have their own lifecycles:

1st image: The management.cattle.io.cluster in the local cluster.
2nd: The provisioning.cattle.io.cluster + ClusterAPI cluster.x-k8s.io.machine machines (and …machineDeployment, but not in this case because custom) in the local cluster.
3rd: The Kubernetes nodes in the downstream cluster.

Things happen to these at different times. The node can be active as far as the downstream k8s is concerned before the machine provisioning process is completed, etc

[gaktive]

Removing the blocker tag from this since this is a known issue of sorts since the UI is simply providing the data given to us. I think there's an existing ticket in rancher/rancher to start ironing out and aligning the views accordingly but local vs. downstream will always be tricky to consoldiate.