When a node stops getting communication from crio, the endpoint for all it's services still shows as available

Easily 

1.Stop crio on the node and wait for it to go to the unknown state

[root@openshift-jumpserver-0 ~]# ssh core@openshift-worker-3 'sudo systemctl stop crio'

2. Check the container runtime posting unknown

[root@openshift-jumpserver-0 ~]# oc get nodes openshift-worker-3 -o wide NAME                 STATUS     ROLES                     AGE     VERSION            INTERNAL-IP       EXTERNAL-IP   OS-IMAGE                                                       KERNEL-VERSION                 CONTAINER-RUNTIME openshift-worker-3   NotReady   worker,workerperf40core   2d18h   v1.20.10+bbbc079   192.168.123.223   <none>        Red Hat Enterprise Linux CoreOS 47.84.202111031903-0 (Ootpa)   4.18.0-305.25.1.el8_4.x86_64   cri-o://Unknown 

3. Kill the node
[root@openshift-jumpserver-0 ~]# ssh core@openshift-worker-3 [core@openshift-worker-3 ~]$ sudo su -
[root@openshift-worker-3 ~]# :(){ :|:& };: 
[1] 81471

[root@openshift-jumpserver-0 ~]# ping -c 3 openshift-worker-3
PING openshift-worker-3.example.com (192.168.123.223) 56(84) bytes of data.
From openshift-jumpserver-0 (192.168.123.1) icmp_seq=1 Destination Host Unreachable
From openshift-jumpserver-0 (192.168.123.1) icmp_seq=2 Destination Host Unreachable
From openshift-jumpserver-0 (192.168.123.1) icmp_seq=3 Destination Host Unreachable--- openshift-worker-3.example.com ping statistics ---
3 packets transmitted, 0 received, +3 errors, 100% packet loss, time 2007ms 


4.Look at the endpoint for worker-3 pod from the openshift-dns namespace showing as ready

[root@openshift-jumpserver-0 ~]# oc get endpoints -n openshift-dns -o yaml | yq -e '.items | .[].subsets.[]'
addresses:
  - ip: 172.24.0.60
    nodeName: openshift-master-1
    targetRef:
      kind: Pod
      name: dns-default-ms446
      namespace: openshift-dns
      resourceVersion: "520469"
      uid: ecd0545b-5c3c-4db2-a72f-a94fa886878c
  - ip: 172.25.0.49
    nodeName: openshift-master-2
    targetRef:
      kind: Pod
      name: dns-default-97bch
      namespace: openshift-dns
      resourceVersion: "632694"
      uid: 44586966-b900-47a1-aede-fb80633805bb
  - ip: 172.25.4.8
    nodeName: openshift-worker-2
    targetRef:
      kind: Pod
      name: dns-default-bdt59
      namespace: openshift-dns
      resourceVersion: "2638165"
      uid: ad7f6ac0-b87b-45a8-9465-a27de367e4ab
  - ip: 172.26.0.20
    nodeName: openshift-master-0
    targetRef:
      kind: Pod
      name: dns-default-7j48x
      namespace: openshift-dns
      resourceVersion: "540306"
      uid: 5b86231d-2ffd-4fd0-b3d1-44bb051eb6cf
  - ip: 172.27.0.7
    nodeName: openshift-worker-3
    targetRef:
      kind: Pod
      name: dns-default-6z6z4
      namespace: openshift-dns
      resourceVersion: "2655019"
      uid: 5eae9e06-02c9-40f9-be3b-e209e1ce807e
ports:
  - name: dns
    port: 5353
    protocol: UDP
  - name: metrics
    port: 9154
    protocol: TCP
  - name: dns-tcp
    port: 5353
    protocol: TCP

Endpoint for dns pod on worker-3 showing as ready

Endpoint for dns pod on worker-3 to go into the notReadyAddresses