Kubeadm init fails

By Cube N Square

Posted October 18, 2021

In Kubernetes

Kubeadm init fails

Problem statement:

In Kubernetes(K8S) 1.22 version, performing K8S installation using kubeadm is resulting in below error.

kubeadm init

Kubelet : Kubelet service isn’t active

Systemctl status kubelet

Check the kubelet log using journalctl command

journalctl -xeu kubelet

Reason

Before we look at the actual reason , lets look what is cgroup

Cgroup : Control groups are used to constrain resources that are allocated to processes. Cgroups are the kernel feature that allows you to set limits for CPU, memory, and disk I/O for one or more processes. By using cgroups, you can isolate a process and the process’s network

Cgroup drivers :

Systemd
Cgroup fs

systemd driver is recommended for kubeadm based setups instead of the cgroupfs driver, because kubeadm manages the kubelet as a systemd service.

Reason for above error is cgroup driver being cgroupfs instead of systemd.

Solution :

Change the cgroup driver to systemd from cgroupfs . Follow below steps

On Master node perform the following steps

sudo vim /etc/docker/daemon.json

{

“exec-opts”: [“native.cgroupdriver=systemd”]

}

### Enter the above three lines in daemon.json and save it ( :wq! ) ###

sudo systemctl daemon-reload

sudo systemctl restart kubelet.service

####reboot the machine before the init command ####

sudo kubeadm init –ignore-preflight-errors=all

# As you would have run kubeadm init command already, running the same command again will through error . To overcome that use the options –ignore-preflight-errors=all

Repeat the same on worker node . But do not run kubeadm init on worker nodes.
Once you have changed the cgroup driver, run the kubeadm join command

Cube N Square