vSphere with Tanzu using NSX-T - Part27 - nullfinalizer kubectl plugin

I have seen many cases where the supervisor namespace gets stuck at Terminating phase waiting on finalization on some of its child resources. This plugin can be used for setting finalizer to null for all objects of a specified api resource under a supervisor namespace. It will be helpful in cleaning up supervisor namespaces stuck terminating phase and can be also used to clean up stale resources under a supervisor namespace.

kubectl-nullfinalizer

#!/bin/bash

Help()
{
   # Display Help
   echo "This plugin sets finalizer to null for specified resource in a namespace."
   echo "Usage: kubectl nullfinalizer SVNAMESPACE RESOURCENAME"
   echo "Example: kubectl nullfinalizer vineetha-svns01 pvc"
}

# Get the options
while getopts ":h" option; do
   case $option in
      h) # display Help
         Help
         exit;;
     \?) # incorrect option
         echo "Error: Invalid option"
         exit;;
   esac
done

kubectl get -n $1 $2 --no-headers | awk '{print $1}' | xargs -I{} kubectl patch -n $1 $2 {} -p '{"metadata":{"finalizers": null}}' --type=merge

Usage

• Place the plugin in the system executable path.
• I placed it in $HOME/.krew/bin in my laptop.
• Once you copied the plugin to the proper path, you can make it executable by: chmod 755 kubectl-nullfinalizer .
• After that you should be able to run the plugin as: kubectl nullfinalizer SUPERVISORNAMESPACE RESOURCENAME .

Example

Following is an exmaple of a supervisor namespace stuck at Terminating phase. While describe you can see that it is waiting on finalization. 

❯ k config current-context
wdc-08-vc07
❯ kg ns svc-sct-bot-dogfooding
NAME                     STATUS        AGE
svc-sct-bot-dogfooding   Terminating   584d
❯
❯ kg ns svc-sct-bot-dogfooding -oyaml

status:
  conditions:
  - lastTransitionTime: "2023-09-26T04:45:21Z"
    message: All resources successfully discovered
    reason: ResourcesDiscovered
    status: "False"
    type: NamespaceDeletionDiscoveryFailure
  - lastTransitionTime: "2023-09-26T04:45:21Z"
    message: All legacy kube types successfully parsed
    reason: ParsedGroupVersions
    status: "False"
    type: NamespaceDeletionGroupVersionParsingFailure
  - lastTransitionTime: "2023-09-26T04:45:21Z"
    message: All content successfully deleted, may be waiting on finalization
    reason: ContentDeleted
    status: "False"
    type: NamespaceDeletionContentFailure
  - lastTransitionTime: "2023-09-26T04:45:21Z"
    message: 'Some resources are remaining: clusters.cluster.x-k8s.io has 1 resource
      instances, kubeadmcontrolplanes.controlplane.cluster.x-k8s.io has 1 resource
      instances, machines.cluster.x-k8s.io has 4 resource instances, persistentvolumeclaims.
      has 9 resource instances, projects.registryagent.vmware.com has 1 resource instances,
      tanzukubernetesclusters.run.tanzu.vmware.com has 1 resource instances'
    reason: SomeResourcesRemain
    status: "True"
    type: NamespaceContentRemaining
  - lastTransitionTime: "2023-09-26T04:45:21Z"
    message: 'Some content in the namespace has finalizers remaining: cluster.cluster.x-k8s.io
      in 1 resource instances, cns.vmware.com/pvc-protection in 9 resource instances,
      controller-finalizer in 1 resource instances, kubeadm.controlplane.cluster.x-k8s.io
      in 1 resource instances, machine.cluster.x-k8s.io in 4 resource instances, tanzukubernetescluster.run.tanzu.vmware.com
      in 1 resource instances'
    reason: SomeFinalizersRemain
    status: "True"
    type: NamespaceFinalizersRemaining
  phase: Terminating

❯ kg pvc -n svc-sct-bot-dogfooding
NAME                                 STATUS        VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS              AGE
gc1-workers-r9jvb-4sfjc-containerd   Terminating   pvc-0d9f4a38-86ad-41d8-ab11-08707780fd85   70Gi       RWO            wdc-08-vc07c01-wcp-mgmt   538d
gc1-workers-r9jvb-szg9r-containerd   Terminating   pvc-ca6b6ec4-85fa-464c-abc6-683358994f3f   70Gi       RWO            wdc-08-vc07c01-wcp-mgmt   538d
gc1-workers-r9jvb-zbdt8-containerd   Terminating   pvc-8f2b0683-ebba-46cb-a691-f79a0e94d0e2   70Gi       RWO            wdc-08-vc07c01-wcp-mgmt   538d
gc2-workers-vpzl2-ffkgx-containerd   Terminating   pvc-69e64099-42c8-44b5-bef2-2737eca49c36   70Gi       RWO            wdc-08-vc07c01-wcp-mgmt   510d
gc2-workers-vpzl2-hww5v-containerd   Terminating   pvc-5a909482-4c95-42c7-b55a-57372f72e75f   70Gi       RWO            wdc-08-vc07c01-wcp-mgmt   510d
gc2-workers-vpzl2-stsnh-containerd   Terminating   pvc-ed7de540-72f4-4832-8439-da471bf4c892   70Gi       RWO            wdc-08-vc07c01-wcp-mgmt   510d
gc3-workers-2qr4c-64xpz-containerd   Terminating   pvc-38478f19-8180-4b9b-b5a9-8c06f17d0fbc   70Gi       RWO            wdc-08-vc07c01-wcp-mgmt   510d
gc3-workers-2qr4c-dpng5-containerd   Terminating   pvc-a8b12657-10bd-4993-b08e-51b7e9b259f9   70Gi       RWO            wdc-08-vc07c01-wcp-mgmt   538d
gc3-workers-2qr4c-wfvvd-containerd   Terminating   pvc-01c6b224-9dc0-4e03-b87e-641d4a4d0d95   70Gi       RWO            wdc-08-vc07c01-wcp-mgmt   538d
❯
❯ k nullfinalizer -h
This plugin sets finalizer to null for specified resource in a namespace.
Usage: kubectl nullfinalizer SVNAMESPACE RESOURCENAME
Example: kubectl nullfinalizer vineetha-svns01 pvc
❯
❯
❯ k nullfinalizer svc-sct-bot-dogfooding pvc
persistentvolumeclaim/gc1-workers-r9jvb-4sfjc-containerd patched
persistentvolumeclaim/gc1-workers-r9jvb-szg9r-containerd patched
persistentvolumeclaim/gc1-workers-r9jvb-zbdt8-containerd patched
persistentvolumeclaim/gc2-workers-vpzl2-ffkgx-containerd patched
persistentvolumeclaim/gc2-workers-vpzl2-hww5v-containerd patched
persistentvolumeclaim/gc2-workers-vpzl2-stsnh-containerd patched
persistentvolumeclaim/gc3-workers-2qr4c-64xpz-containerd patched
persistentvolumeclaim/gc3-workers-2qr4c-dpng5-containerd patched
persistentvolumeclaim/gc3-workers-2qr4c-wfvvd-containerd patched
❯
❯
❯ kg projects.registryagent.vmware.com -n svc-sct-bot-dogfooding
NAME                     AGE
svc-sct-bot-dogfooding   584d
❯
❯ k nullfinalizer -h
This plugin sets finalizer to null for specified resource in a namespace.
Usage: kubectl nullfinalizer SVNAMESPACE RESOURCENAME
Example: kubectl nullfinalizer vineetha-svns01 pvc
❯
❯ k nullfinalizer svc-sct-bot-dogfooding projects.registryagent.vmware.com
project.registryagent.vmware.com/svc-sct-bot-dogfooding patched
❯
❯
❯ kg ns svc-sct-bot-dogfooding
Error from server (NotFound): namespaces "svc-sct-bot-dogfooding" not found
❯
 

Hope it was useful. Cheers!

vSphere with Tanzu using NSX-T - Part26 - Jumpbox kubectl plugin to SSH to TKC node

For troubleshooting TKC (Tanzu Kubernetes Cluster) you may need to ssh into the TKC nodes. For doing ssh, you will need to first create a jumpbox pod under the supervisor namespace and from there you can ssh to the TKC nodes.

Here is the manual procedure: https://docs.vmware.com/en/VMware-vSphere/7.0/vmware-vsphere-with-tanzu/GUID-587E2181-199A-422A-ABBC-0A9456A70074.html

Following kubectl plugin creats a jumpbox pod under a supervisor namespace. You can exec into this jumpbox pod to ssh into the TKC VMs.

kubectl-jumpbox

#!/bin/bash

Help()
{
   # Display Help
   echo "Description: This plugin creats a jumpbox pod under a supervisor namespace. You can exec into this jumpbox pod to ssh into the TKC VMs."
   echo "Usage: kubectl jumpbox SVNAMESPACE TKCNAME"
   echo "Example: k exec -it jumpbox-tkc1 -n svns1 -- /usr/bin/ssh vmware-system-user@VMIP"
}

# Get the options
while getopts ":h" option; do
   case $option in
      h) # display Help
         Help
         exit;;
     \?) # incorrect option
         echo "Error: Invalid option"
         exit;;
   esac
done

kubectl create -f - <<EOF
apiVersion: v1
kind: Pod
metadata:
  name: jumpbox-$2
  namespace: $1           #REPLACE
spec:
  containers:
  - image: "photon:3.0"
    name: jumpbox
    command: [ "/bin/bash", "-c", "--" ]
    args: [ "yum install -y openssh-server; mkdir /root/.ssh; cp /root/ssh/ssh-privatekey /root/.ssh/id_rsa; chmod 600 /root/.ssh/id_rsa; while true; do sleep 30; done;" ]
    volumeMounts:
      - mountPath: "/root/ssh"
        name: ssh-key
        readOnly: true
    resources:
      requests:
        memory: 2Gi
  
  volumes:
    - name: ssh-key
      secret:
        secretName: $2-ssh     #REPLACE YOUR-CLUSTER-NAME-ssh 

  
EOF

Usage

• Place the plugin in the system executable path.
• I placed it in $HOME/.krew/bin directory in my laptop.
• Once you copied the plugin to the proper path, you can make it executable by: chmod 755 kubectl-jumpbox
• After that you should be able to run the plugin as: kubectl jumpbox SUPERVISORNAMESPACE TKCNAME

 

Example

❯ kg tkc -n vineetha-dns1-test
NAME               CONTROL PLANE   WORKER   TKR NAME                           AGE    READY   TKR COMPATIBLE   UPDATES AVAILABLE
tkc                1               3        v1.21.6---vmware.1-tkg.1.b3d708a   213d   True    True             [1.22.9+vmware.1-tkg.1.cc71bc8]
tkc-using-cci-ui   1               1        v1.23.8---vmware.3-tkg.1           37d    True    True
❯
❯ kg po -n vineetha-dns1-test
NAME         READY   STATUS    RESTARTS   AGE
nginx-test   1/1     Running   0          29d
❯
❯
❯ kubectl jumpbox vineetha-dns1-test tkc
pod/jumpbox-tkc created
❯
❯ kg po -n vineetha-dns1-test
NAME          READY   STATUS    RESTARTS   AGE
jumpbox-tkc   0/1     Pending   0          8s
nginx-test    1/1     Running   0          29d
❯
❯ kg po -n vineetha-dns1-test
NAME          READY   STATUS    RESTARTS   AGE
jumpbox-tkc   1/1     Running   0          21s
nginx-test    1/1     Running   0          29d
❯
❯ k jumpbox -h
Description: This plugin creats a jumpbox pod under a supervisor namespace. You can exec into this jumpbox pod to ssh into the TKC VMs.
Usage: kubectl jumpbox SVNAMESPACE TKCNAME
Example: k exec -it jumpbox-tkc1 -n svns1 -- /usr/bin/ssh vmware-system-user@VMIP
❯
❯ kg vm -n vineetha-dns1-test -o wide
NAME                                                              POWERSTATE   CLASS               IMAGE                                                       PRIMARY-IP      AGE
tkc-control-plane-8rwpk                                           poweredOn    best-effort-small   ob-18900476-photon-3-k8s-v1.21.6---vmware.1-tkg.1.b3d708a   172.29.0.7      133d
tkc-using-cci-ui-control-plane-z8fkt                              poweredOn    best-effort-small   ob-20953521-tkgs-ova-photon-3-v1.23.8---vmware.3-tkg.1      172.29.13.130   37d
tkc-using-cci-ui-tkg-cluster-nodepool-9nf6-n6nt5-b97c86fb45mvgj   poweredOn    best-effort-small   ob-20953521-tkgs-ova-photon-3-v1.23.8---vmware.3-tkg.1      172.29.13.131   37d
tkc-workers-zbrnv-6c98dd84f9-52gn6                                poweredOn    best-effort-small   ob-18900476-photon-3-k8s-v1.21.6---vmware.1-tkg.1.b3d708a   172.29.0.6      133d
tkc-workers-zbrnv-6c98dd84f9-d9mm7                                poweredOn    best-effort-small   ob-18900476-photon-3-k8s-v1.21.6---vmware.1-tkg.1.b3d708a   172.29.0.8      133d
tkc-workers-zbrnv-6c98dd84f9-kk2dg                                poweredOn    best-effort-small   ob-18900476-photon-3-k8s-v1.21.6---vmware.1-tkg.1.b3d708a   172.29.0.3      133d
❯
❯ k exec -it jumpbox-tkc -n vineetha-dns1-test -- /usr/bin/ssh vmware-system-user@172.29.0.7
The authenticity of host '172.29.0.7 (172.29.0.7)' can't be established.
ECDSA key fingerprint is SHA256:B7ptmYm617lFzLErJm7G5IdT7y4SJYKhX/OenSgguv8.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added '172.29.0.7' (ECDSA) to the list of known hosts.
Welcome to Photon 3.0 (\m) - Kernel \r (\l)
 13:06:06 up 133 days,  4:46,  0 users,  load average: 0.23, 0.33, 0.27

36 Security notice(s)
Run 'tdnf updateinfo info' to see the details.
vmware-system-user@tkc-control-plane-8rwpk [ ~ ]$ sudo su
root [ /home/vmware-system-user ]#
root [ /home/vmware-system-user ]#

Hope it was useful. Cheers!