Dashboard v2 import cluster has 0 machines displayed on the UI after the nodes are active

[anupama2501]

master-6030ab58d56bbc10ca4764c8c3438fcbfb041206-head

1. Create a kubernetes cluster using rke
2. Import the cluster into the rancher server with master-head commit 6030ab by running the kubectl in the nodes
3. Once the cluster is active --> verify from the cluster management --> imported cluster shows 0 machines
4. The nodes can be seen active from the explore cluster --> imported cluster --> cluster --> nodes
5. Enhancement to show the number of nodes registered under machines and the nodes to be accessible from cluster management for the imported cluster is needed

[vincent99]

I think we should probably do something to clarify this, but this isn't really a bug.. "Provisioning" clusters and nodes/machines are separate from the "Management" cluster and the k8s Node type.

When you're inside the cluster (explorer), you're looking at the k8s node resources, which exist no matter how the cluster was deployed.

When you're in cluster management, you're looking at the resources we actually provisioned and manage. There are no nodes shown here because we didn't provision any nodes for those clusters. If you delete the cluster, no 'physical' infrastructure will be deleted. If you make an RKE2+EC2 cluster in there, there will be nodes because we deployed them to EC2, and will clean them up when you delete the cluster.

This is also why they're (sometimes, inconsistently) called "Machines" versus "Nodes".

[izaac]

Rancher version 2.6 master-head 55f0bb2

I am unable to see the Machines we provisioned with RKE1 from Cluster Explorer -> Cluster Management I understand the imported clusters issue but shouldn't we display these Machines/Nodes here for RKE1?

The cluster information shows the Nodes total of the RKE1 cluster I provisioned

[izaac]

I've opened an issue for RKE1 specifically, this is for imported clusters.

[nwmac]

As far as I can tell, this is working correctly - moving to test for QA to validate

[jameson-mcghee]

I attempted to reproduce this issue on both v2.6.7 and v2.6-head (Commit ID: 400e559), and I was unable to reproduce this issue on either version. Based on that fact, @nwmac's previous comment, and confirmation from both @anupama2501 and @Josh-Diamond, the decision was made to close this ticket as it is unreproducible.

Repro Steps: (Followed exact steps laid out in the ticket's Description)

v2.6.7