cluster controller fail to get cluster status

[mfanjie]

When I create cluster with last master branch, by post request to federation apiserver, I got the following error logs

E0528 23:51:29.446498   42339 runtime.go:58] Recovered from panic: &runtime.TypeAssertionError{interfaceString:"interface {}", concreteString:"*federation.Cluster", assertedString:"*v1alpha1.Cluster", missingMethod:""} (interface conversion: interface {} is *federation.Cluster, not *v1alpha1.Cluster)
/Users/fmeng/code/kubernetes/src/k8s.io/kubernetes/pkg/util/runtime/runtime.go:52
/Users/fmeng/code/kubernetes/src/k8s.io/kubernetes/pkg/util/runtime/runtime.go:40
/usr/local/go/src/runtime/asm_amd64.s:472
/usr/local/go/src/runtime/panic.go:443
/usr/local/go/src/runtime/iface.go:182
/Users/fmeng/code/kubernetes/src/k8s.io/kubernetes/federation/pkg/federation-controller/cluster/clustercontroller.go:97
/Users/fmeng/code/kubernetes/src/k8s.io/kubernetes/federation/pkg/federation-controller/cluster/clustercontroller.go:80
/Users/fmeng/code/kubernetes/src/k8s.io/kubernetes/pkg/controller/framework/controller.go:170
<autogenerated>:25
/Users/fmeng/code/kubernetes/src/k8s.io/kubernetes/pkg/controller/framework/controller.go:248
/Users/fmeng/code/kubernetes/src/k8s.io/kubernetes/pkg/controller/framework/controller.go:128
/Users/fmeng/code/kubernetes/src/k8s.io/kubernetes/pkg/controller/framework/controller.go:103
/Users/fmeng/code/kubernetes/src/k8s.io/kubernetes/pkg/util/wait/wait.go:86
/Users/fmeng/code/kubernetes/src/k8s.io/kubernetes/pkg/util/wait/wait.go:87
/Users/fmeng/code/kubernetes/src/k8s.io/kubernetes/pkg/util/wait/wait.go:49
/Users/fmeng/code/kubernetes/src/k8s.io/kubernetes/pkg/controller/framework/controller.go:103
/usr/local/go/src/runtime/asm_amd64.s:1998

and then the following error was dumped repeatedly

I0528 23:52:02.142519   42339 clustercontroller.go:127] Failed to create cluster client, err: unexpected: POD_NAMESPACE env var returned empty string
I0528 23:52:02.142525   42339 clustercontroller.go:165] Failed to Get the status of cluster: cluster179

something wrong in clustercontroller.go make the controller manager failed to get the latest cluster status, this impact other controllers like service controller which depends on the status of clusters.

The workaround is to manually call cluster status update api to update the cluster to ready, then service controller will work properly.
example:
sent put request to the cluster status api like localhost:8080/apis/federation/v1alpha1/clusters/cluster179/status
with the following body.

{
  "kind": "Cluster",
  "apiVersion": "federation/v1alpha1",
  "metadata": {
    "name": "cluster179"
  },
  "status": {
    "conditions": [{
        "type": "Ready",
        "status": "True"
    }]
  }
}

[mfanjie]

@huangyuqi

[mfanjie]

fixed by #26550