Setup containerd for K8S on Alpine Linux

setup_containerd_for_k8s.sh

#!/bin/bash
set -e

## CRI containrd
## CNI flannel

# Alpine Edge only! Let's hope alpine 3.13 or 3.14 would have k8s in main tree
echo http://dl-cdn.alpinelinux.org/alpine/edge/testing >> /etc/apk/repositories
echo http://dl-cdn.alpinelinux.org/alpine/edge/community >> /etc/apk/repositories

# Install containerd
apk add containerd containerd-openrc

# Load required modules
cat > /etc/modules-load.d/containerd.conf <<EOF
overlay
br_netfilter
EOF

modprobe overlay
modprobe br_netfilter

# Setup required sysctl params, these persist across reboots.
cat > /etc/sysctl.d/99-kubernetes-cri.conf <<EOF
net.bridge.bridge-nf-call-iptables  = 1
net.ipv4.ip_forward                 = 1
net.bridge.bridge-nf-call-ip6tables = 1
EOF

sysctl -p
sysctl -p /etc/sysctl.d/99-kubernetes-cri.conf

# Create default containerd config
mkdir -p /etc/containerd

# See https://git.alpinelinux.org/aports/commit/?id=72a355e1c8437c4e32a3e22bc3888905a6e545ba for details on why we need update the plugin dir
containerd config default | sed "s|/opt/cni/bin|/usr/libexec/cni|g" > /etc/containerd/config.toml

rc-update add containerd default
rc-service containerd start


# K8S - the main part!
apk add kubernetes conntrack-tools kubeadm kubectl kubelet cri-tools

rc-update add kubelet default

# Prevent conflicts between docker iptables (packet filtering) rules and k8s pod communication
# See https://github.com/kubernetes/kubernetes/issues/40182 for further details.
iptables -P FORWARD ACCEPT

# disable swap
sed -i '/swap/d' /etc/fstab
swapoff -a


# set up cluster
echo "172.42.42.100   kmaster kmaster.example.com" >> /etc/hosts
echo "172.42.42.101   kworker1 kworker1.example.com" >> /etc/hosts
echo "172.42.42.102   kmaster2 kworker2.example.com" >> /etc/hosts


## set up master node - DO NOT do this step for worker node!!
kubeadm config images pull

## start 2 sessions. The 1st session we will try to kickstart the cluster
kubeadm init --cri-socket /run/containerd/containerd.sock --apiserver-advertise-address=172.42.42.100 --pod-network-cidr=192.168.0.0/16 --v=5

## once we get
## [wait-control-plane] Waiting for the kubelet to boot up the control plane as static Pods from directory "/etc/kubernetes/manifests". This can take up to 4m0s
## we will kickstart kubelet daemon in the 2nd session
rc-service kubelet start

# Copy Kube admin config
echo "Copy kube admin config to user .kube directory"

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

# verify that we have our master node

kubectl get nodes

# please save the output of join command for the worker node
kubeadm token create --print-join-command > join.sh

## Next let's setup our worker node, simply repeat the same step to get kubelet running first, then run the join.sh command