Kubernetes (K8s) Cluster setup - Ubuntu 22.04 : Network plugin returns error: cni plugin not initialized

General Discussions
,
1

Hi Team,

Currently facing some issues with Kubernetes clusters (1 master node, 1 worker node) setup in Ubuntu 22.04.

I am following exact steps mentioned in this [blog] (How to Install Kubernetes (K8s) on Ubuntu 24.04 - HostnExtra) to setup k8s cluster.

Step 1: Update and Upgrade the System (all nodes)

sudo apt update && sudo apt upgrade -y

Step 2: Install Docker (all nodes)

sudo apt install -y docker.io

Note: Docker is installed and status is active.

Step 3: Install Kubernetes Components (all nodes)

curl -fsSL https://pkgs.k8s.io/core:/stable:/v1.30/deb/Release.key | sudo gpg --dearmor -o /etc/apt/keyrings/kubernetes-apt-keyring.gpg
echo "deb [signed-by=/etc/apt/keyrings/kubernetes-apt-keyring.gpg] 

https://pkgs.k8s.io/core:/stable:/v1.30/deb/ /" | sudo tee /etc/apt/sources.list.d/kubernetes.list

Install the Kubernetes components:

sudo apt update
sudo apt install -y kubelet kubeadm kubectl
sudo apt-mark hold kubelet kubeadm kubectl

Step 4: Disable Swap (all nodes)

sudo swapoff -a
sudo sed -i '/ swap / s/^\(.*\)$/#\1/g' /etc/fstab

Step 5: Initialize the Master Node (Master node)

sudo kubeadm init --pod-network-cidr=10.244.0.0/16

Output:

kubeadm join 10.x.x.x:6443 --token z8gd9n.abccd \
        --discovery-token-ca-cert-hash sha256:9967fc499fca61705d206axxx

Step 6: Configure kubectl for the Master Node (Master node - Non root user)

mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config

Step 7: Install a Pod Network Add-on on Master node (Master node)

kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml

Step 8: Join Worker Nodes to the Cluster (Worker node)

kubeadm join 10.x.x.x:6443 --token z8gd9n.abccd \
        --discovery-token-ca-cert-hash sha256:9967fc499fca61705d206axxx

Step 9: verifying cluster

kubectl get nodes

Output:

NAME       STATUS     ROLES           AGE   VERSION
pocmst      NotReady   control-plane   86m   v1.30.6
pocwrk      NotReady   <none>          82m   v1.30.6

Note: both nodes are NotReady state.

In order to troubleshoot this Notready state, followed below procedure.

Step 1: kubectl describe node pocmst

Conditions:
  Type             Status  LastHeartbeatTime                 LastTransitionTime                Reason                       Message
  ----             ------  -----------------                 ------------------                ------                       -------
  MemoryPressure   False   Tue, 19 Nov 2024 13:20:26 +0000   Tue, 19 Nov 2024 11:48:04 +0000   KubeletHasSufficientMemory   kubelet has sufficient memory available
  DiskPressure     False   Tue, 19 Nov 2024 13:20:26 +0000   Tue, 19 Nov 2024 11:48:04 +0000   KubeletHasNoDiskPressure     kubelet has no disk pressure
  PIDPressure      False   Tue, 19 Nov 2024 13:20:26 +0000   Tue, 19 Nov 2024 11:48:04 +0000   KubeletHasSufficientPID      kubelet has sufficient PID available
  Ready            False   Tue, 19 Nov 2024 13:20:26 +0000   Tue, 19 Nov 2024 11:48:04 +0000   KubeletNotReady              container runtime network not ready: NetworkReady=false reason:NetworkPluginNotReady message:Network plugin returns error: cni plugin not initialized

Noted the issue: container runtime network not ready: NetworkReady=false reason:NetworkPluginNotReady message:Network plugin returns error: cni plugin not initialized

Step 2: journalctl -u kubelet

 E1119 13:27:42.392914    2139 pod_workers.go:1298] "Error syncing pod, skipping" err="failed to \"StartContainer\" for \"install-cni-plugin\" with ImagePullBackOff: \"Back-off pulling image \\\"docker.io/flannel/flannel-cni-plugin:v1.6.0-flannel1\\\"\"" pod="kube-flannel/kube-flannel-ds-hlsgr" podUID="b3d8f1d3-2a77-448f-b2df-3ee5ebe629d4"
Nov 19 13:27:46 wso2dmigpocmstr kubelet[2139]: E1119 13:27:46.078441    2139 kubelet.go:2920] "Container runtime network not ready" networkReady="NetworkReady=false reason:NetworkPluginNotReady message:Network plugin returns error: cni plugin not initialized"

Step 3:

kubectl get pod -A

NAMESPACE      NAME                                      READY   STATUS                  RESTARTS       AGE
kube-flannel   kube-flannel-ds-hlsgr                     0/1     Init:ImagePullBackOff   0              100m
kube-flannel   kube-flannel-ds-mvdc2                     0/1     Init:ImagePullBackOff   0              98m
kube-system    coredns-55cb58b774-7l6gm                  0/1     Pending                 0              102m
kube-system    coredns-55cb58b774-7lxtq                  0/1     Pending                 0              102m
kube-system    etcd-pocmstr                              1/1     Running                 39 (54m ago)   102m
kube-system    kube-apiserver-pocmstr                  1/1     Running                 1 (54m ago)    102m
kube-system    kube-controller-manager-pocmstr           1/1     Running                 11 (54m ago)   102m
kube-system    kube-proxy-k5x2x                          1/1     Running                 1 (54m ago)    102m
kube-system    kube-proxy-w2qwz                          0/1     ImagePullBackOff        0              98m

Is there any steps am i missing while setting up k8s cluster? May i know any suggestion to resolve this issue? so that All nodes in cluster will become as READY state.

2

flannel is network cni, from the pod status above, the images pull fail in the flannel pods .

  1. you can check the image registry correct or not: kubectl describe pods kube-flannel-ds-hlsgr -n kube-system
    and your can try to download mannually(image registry get from kubectl describe): docker pull image-registry
  2. check the PROXY and NO_PROXY envriment variable, make sure the pull images request can forward to correct network
    export HTTP_PROXY=…
    export HTTPS_PROXY=…
    export NO_PROXY=…
  3. your coredns pods is pending, you should check too by kubectl describe