When filtering etcd or api-server pods, the readiness status of the pods was not checked.

[yxxhero]

What keywords did you search in kubeadm issues before filing this one?

If you have found any duplicates, you should instead reply there and close this page.

If you have not found any duplicates, delete this section and continue on.

Is this a BUG REPORT or FEATURE REQUEST?

Choose one: BUG REPORT or FEATURE REQUEST
BUG

Versions

kubeadm version (use kubeadm version): all

Environment:

• Kubernetes version (use kubectl version):
• Cloud provider or hardware configuration:
• OS (e.g. from /etc/os-release):
• Kernel (e.g. uname -a):
• Container runtime (CRI) (e.g. containerd, cri-o):
• Container networking plugin (CNI) (e.g. Calico, Cilium):
• Others:

What happened?

When filtering etcd or api-server pods, the readiness status of the pods was not checked.

What you expected to happen?

When filtering etcd or api-server pods, the readiness status of the pods was checked.

How to reproduce it (as minimally and precisely as possible)?

Anything else we need to know?

[Lubomir I. Ivanov]

you need to provide more information.

[yxxhero]

func getRawEtcdEndpointsFromPodAnnotationWithoutRetry(client clientset.Interface) ([]string, int, error) {
	klog.V(3).Infof("retrieving etcd endpoints from %q annotation in etcd Pods", constants.EtcdAdvertiseClientUrlsAnnotationKey)
	podList, err := client.CoreV1().Pods(metav1.NamespaceSystem).List(
		context.TODO(),
		metav1.ListOptions{
			LabelSelector: fmt.Sprintf("component=%s,tier=%s", constants.Etcd, constants.ControlPlaneTier),
		},
	)
	if err != nil {
		return []string{}, 0, err
	}
	etcdEndpoints := []string{}
	for _, pod := range podList.Items {
		etcdEndpoint, ok := pod.ObjectMeta.Annotations[constants.EtcdAdvertiseClientUrlsAnnotationKey]
		if !ok {
			klog.V(3).Infof("etcd Pod %q is missing the %q annotation; cannot infer etcd advertise client URL using the Pod annotation", pod.ObjectMeta.Name, constants.EtcdAdvertiseClientUrlsAnnotationKey)
			continue
		}
		etcdEndpoints = append(etcdEndpoints, etcdEndpoint)
	}
	return etcdEndpoints, len(podList.Items), nil
}

we can see the code above. it doesn't check the ready status of pod. @neolit123

[Lubomir I. Ivanov]

we can see the code above. it doesn't check the ready status of pod. @neolit123

why is it a problem that we need to fix?

[yxxhero]

etcd-node001                                      0/1     CrashLoopBackOff    18074 (3m20s ago)   4d16h   1.1.1.1        node001     <none>           <none>
etcd-node002                                      1/1     Running             12 (6d3h ago)       69d     1.1.1.1       node002     <none>           <none>
etcd-node005                                      1/1     Running             10 (8d ago)         69d     1.1.1.1       node005     <none>           <none>
etcd-node003                                      1/1     Running             8 (19d ago)         67d     1.1.1.1       node003     <none>           <none>
etcd-node004                                      1/1     Running             7 (60d ago)         67d     1.1.1.1        node004     <none>           <none>

see the above. shall we add the unhealth pod as etcd node? @neolit123 thanks so much.

[Lubomir I. Ivanov]

i think getRawEtcdEndpointsFromPodAnnotationWithoutRetry is supposed to get the annotations even if the pod is not ready. readyness is checked during upgrade.

are you seeing this as an actual problem on your clusters?

[yxxhero]

@neolit123 make sense. maybe we can log something to nitice user?

[Lubomir I. Ivanov]

we could log with klog.V(3).Infof if the Pod is not ready. to my understanding this is what you'd like to be changed?

[yxxhero]

Yeah. It's ok for now. I will do that. Alright？ Thanks So much @neolit123