Nothing happens if I try to set a non-responsive cluster as current cluster

[adietish]

related to

• Kubernetes by Red Hat (1.5.0.8) locks up entire editor for 30+ seconds #862
• When I use a non-responsive cluster, the IDE is freezing when I close an editor that contains a k8 yaml resource #869

---

Steps:

1. ASSERT: have a cluster that's shut down (ex. have kind cluster defined in kubeconfig, but it's not running)
2. EXEC: in resource tree: right click on the entry for the non-responsive cluster and pick "Set as current cluster"

Result:
Nothing happens. In the status bar you see a running operation "Setting kind as current cluster". This operation can easily be running for several minutes.
Even worse is that the IDE cannot be closed gracefully when this operation is running. You have to forcefully kill it.

---

Steps 2:

1. ASSERT: have a cluster that's shut down (ex. have kind cluster defined in kubeconfig, but it's not running)
2. EXEC: start up the IDE, open the Kubernetes tool window. Wait for the contexts to be listed.

Result:
It takes several seconds (~20s in my case) until the contexts are listed

---

I suspect this to have been introduced with the switch to the new kubernetes-client 7.0.0

[adietish]

The culprit is the check if the new current cluster is an OpenShift cluster. It's executed on any cluster (so that either a kubernetes- or an OpenShift-client is created). This check runs into a timeout.