Comment 8 for bug 1867398

Deployed the following bundle: http://paste.ubuntu.com/p/tdjqQ3GjJ2/

Followed the reproducer steps.

### With current bionic-updates version 1.3.3-0ubuntu1~18.04.1, problem reproduced. #####

ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ kubectl delete pod --all
pod "busybox" deleted
ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ juju run --application kubernetes-worker "sudo grep -i niedbalski /etc/containerd/* | grep -i endpoint"
- Stdout: |
    /etc/containerd/config.toml: endpoint = ["niedbalski-bastion.cloud.sts:5000"]
  UnitId: kubernetes-worker/0
- ReturnCode: 1
  Stdout: ""
  UnitId: kubernetes-worker/1

ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ juju run --application kubernetes-worker "sudo dpkg -l |grep containerd"
- Stdout: |
    ii containerd 1.3.3-0ubuntu1~18.04.1 amd64 daemon to control runC
  UnitId: kubernetes-worker/0

ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ kubectl delete pod --all
pod "busybox" deleted
ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ kubectl get nodes
NAME STATUS ROLES AGE VERSION
juju-3a79d2-00268738-4 Ready <none> 13d v1.16.8
juju-3a79d2-00268738-5 Ready,SchedulingDisabled <none> 13d v1.16.8
ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ kubectl get pod -o wide
No resources found in default namespace.
(reverse-i-search)`deplo': kubectl edit ^Cployment microbot
(reverse-i-search)`appl': juju run --^Cplication kubernetes-worker "sudo dpkg -l |grep containerd"

ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ kubectl apply -f busybox.yaml
pod/busybox created
ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ kubectl get pod -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
busybox 0/1 ErrImagePull 0 3s 10.1.84.4 juju-3a79d2-00268738-4 <none> <none>

### With current -proposed version 1.3.3-0ubuntu1~18.04.2, problem is fixed. #####

ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ kubectl get nodes
NAME STATUS ROLES AGE VERSION
juju-3a79d2-00268738-4 Ready <none> 13d v1.16.8
juju-3a79d2-00268738-5 Ready,SchedulingDisabled <none> 13d v1.16.8
ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ kubectl get pod -o wide
No resources found in default namespace.
ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ kubectl apply -f busybox.yaml
pod/busybox created
ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ kubectl get nodes
NAME STATUS ROLES AGE VERSION
juju-3a79d2-00268738-4 Ready <none> 13d v1.16.8
juju-3a79d2-00268738-5 Ready,SchedulingDisabled <none> 13d v1.16.8
ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ kubectl get pod -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
busybox 1/1 Running 0 6s 10.1.84.3 juju-3a79d2-00268738-4 <none> <none>

ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ juju run --application kubernetes-worker "sudo dpkg -l |grep containerd"
- Stdout: |
    ii containerd 1.3.3-0ubuntu1~18.04.2 amd64 daemon to control runC
  UnitId: kubernetes-worker/0
- Stdout: |
    ii containerd 1.3.3-0ubuntu1~18.04.1 amd64 daemon to control runC
  UnitId: kubernetes-worker/1

ubuntu@niedbalski-bastion:~/stsstack-bundles/kubernetes/00268738$ more busybox.yaml
apiVersion: v1
kind: Pod
metadata:
  name: busybox
  namespace: default
spec:
  containers:
    - name: busybox
      image: niedbalski-bastion.cloud.sts:5000/busybox:latest
      command:
        - sleep
        - "3600"
  imagePullSecrets:
    - name: regcred
  restartPolicy: Always

Therefore the verification is done.