Deploy k3s(kubernetes) ha cluster with 3 master nodes

Deploy k3s(kubernetes) ha cluster with 3 master node with etcd storage

Why k3s ? Read there

Req:

• ubuntu 20.04

• 6 nodes(you can use 2 master nodes for make HA Cluster with mysql or postgres database as cluster backend)

WAY1 – Using etcd as database backend

In K3s v1.19.1+, embedded etcd is used.

To run K3s in this mode, you must have an odd number of server nodes. We recommend starting with three nodes.

Prepare file /etc/hosts on all nodes cluster.

192.168.122.104 k3s-master-1
192.168.122.105 k3s-master-2
192.168.122.106 k3s-master-3
192.168.122.101 k3s-worker-1
192.168.122.102 k3s-worker-2
192.168.122.103 k3s-worker-3

MASTER 1 INIT

Install k3s with official way with latest version(1.19.3)

export K3S_KUBECONFIG_MODE="644"
export INSTALL_K3S_CHANNEL="latest"
curl -sfL https://get.k3s.io |  sh -

get token for k3s

cat /var/lib/rancher/k3s/server/node-token

kill porcess on port 6443

kill -9 $(lsof -i :6443 | grep LISTEN | awk '{print $2}')

And make cluster init

K3S_TOKEN="K10da6d3b43164285d34f09aa5c38464c94ab7b5254381347cad1388430934d58e1::server:0e14477cea9cf607c150d6e35425274e" k3s server --cluster-init

wait minute and press ctrl+c. Skip errors

Restart k3s service via systemctl

systemctl restart k3s

MASTER 2,3 ADD to Master1

Install k3s

export K3S_KUBECONFIG_MODE="644"
export INSTALL_K3S_CHANNEL="latest"
curl -sfL https://get.k3s.io |  sh -

Connect to exist masters

kill -9 $(lsof -i :6443 | grep LISTEN | awk '{print $2}')
K3S_TOKEN="K10da6d3b43164285d34f09aa5c38464c94ab7b5254381347cad1388430934d58e1::server:0e14477cea9cf607c150d6e35425274e" k3s server --server https://k3s-master-1:6443

wait minute and press ctrl+c. Skip errors

Restart k3s service via systemctl

systemctl restart k3s

Done

WORKERS ADD

export K3S_KUBECONFIG_MODE="644"
export INSTALL_K3S_CHANNEL="latest"
export K3S_URL="https://k3s-master-1:6443"
export K3S_TOKEN="K10da6d3b43164285d34f09aa5c38464c94ab7b5254381347cad1388430934d58e1::server:0e14477cea9cf607c150d6e35425274e"
curl -sfL https://get.k3s.io |  sh -

SEE CLUSTER STATUS

root@vm104:~# kubectl get node

NAME    STATUS     ROLES         AGE     VERSION
vm101   Ready      worker        4m11s   v1.19.3+k3s2
vm102   Ready      worker        3m55s   v1.19.3+k3s2
vm103   Ready      worker        3m54s   v1.19.3+k3s2
vm104   Ready      etcd,master   15m     v1.19.3+k3s2
vm105   NotReady   etcd,master   9m44s   v1.19.3+k3s2
vm106   Ready      etcd,master   6m16s   v1.19.3+k3s2

PROFIT!