Building a basic kubernetes cluster with kubeadm
This is just a basic setup installing a 3 node Kubernetes setup on 3 nodes. The nodes can probably be anywhere, AWS, GCP, VMware etc as long as they are running Ubuntu 20.04.
One node needs to be designated as the master and the other 2 as workers.
To start with the pre-requisites need to be installed on all 3 nodes. This can be done via a script or by entering the commands below individually
Download and run the script
1curl https://raw.githubusercontent.com/narmitag/kubernetes/main/basic_setup/setup_machine.sh -o setup_machine.sh
2chmod +x setup_machine.sh
3sudo ./setup_machine.sh
Running the steps:
1#Install containerd
2cat <<EOF | sudo tee /etc/modules-load.d/containerd.conf
3overlay
4br_netfilter
5EOF
6
7sudo modprobe overlay
8sudo modprobe br_netfilter
9
10cat <<EOF | sudo tee /etc/sysctl.d/99-kubernetes-cri.conf
11net.bridge.bridge-nf-call-iptables = 1
12net.ipv4.ip_forward = 1
13net.bridge.bridge-nf-call-ip6tables = 1
14EOF
15
16sudo sysctl --system
17sudo apt-get update && sudo apt-get install -y containerd
18
19sudo mkdir -p /etc/containerd
20sudo containerd config default | sudo tee /etc/containerd/config.toml
21sudo systemctl restart containerd
22
23#Disable swap
24sudo swapoff -a
25sudo sed -e '/swap/ s/^#*/#/' -i /etc/fstab
26curl -s https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add -
27
28#Install Kubernetes binaries
29cat <<EOF | sudo tee /etc/apt/sources.list.d/kubernetes.list
30deb https://apt.kubernetes.io/ kubernetes-xenial main
31EOF
32
33sudo apt-get update
34sudo apt-get install -y kubelet=1.24.0-00 kubeadm=1.24.0-00 kubectl=1.24.0-00
35sudo apt-mark hold kubelet kubeadm kubectl
The cluster now needs to be bootstrapped from the master
1$ sudo kubeadm init --pod-network-cidr 192.168.0.0/16 --kubernetes-version 1.24.0
2
3$ mkdir -p $HOME/.kube
4$ sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
5$ sudo chown $(id -u):$(id -g) $HOME/.kube/config
6
7$ kubectl get no
8NAME STATUS ROLES AGE VERSION
9master NotReady control-plane 50s v1.24.0
A network plugin needs to be installed to get the node ready (it may take a few minutes for the node to become ready). This will install calico
1$ kubectl apply -f https://docs.projectcalico.org/manifests/calico.yaml
2
3poddisruptionbudget.policy/calico-kube-controllers created
4serviceaccount/calico-kube-controllers created
5serviceaccount/calico-node created
6configmap/calico-config created
7customresourcedefinition.apiextensions.k8s.io/bgpconfigurations.crd.projectcalico.org created
8customresourcedefinition.apiextensions.k8s.io/bgppeers.crd.projectcalico.org created
9customresourcedefinition.apiextensions.k8s.io/blockaffinities.crd.projectcalico.org created
10customresourcedefinition.apiextensions.k8s.io/caliconodestatuses.crd.projectcalico.org created
11customresourcedefinition.apiextensions.k8s.io/clusterinformations.crd.projectcalico.org created
12customresourcedefinition.apiextensions.k8s.io/felixconfigurations.crd.projectcalico.org created
13customresourcedefinition.apiextensions.k8s.io/globalnetworkpolicies.crd.projectcalico.org created
14customresourcedefinition.apiextensions.k8s.io/globalnetworksets.crd.projectcalico.org created
15customresourcedefinition.apiextensions.k8s.io/hostendpoints.crd.projectcalico.org created
16customresourcedefinition.apiextensions.k8s.io/ipamblocks.crd.projectcalico.org created
17customresourcedefinition.apiextensions.k8s.io/ipamconfigs.crd.projectcalico.org created
18customresourcedefinition.apiextensions.k8s.io/ipamhandles.crd.projectcalico.org created
19customresourcedefinition.apiextensions.k8s.io/ippools.crd.projectcalico.org created
20customresourcedefinition.apiextensions.k8s.io/ipreservations.crd.projectcalico.org created
21customresourcedefinition.apiextensions.k8s.io/kubecontrollersconfigurations.crd.projectcalico.org created
22customresourcedefinition.apiextensions.k8s.io/networkpolicies.crd.projectcalico.org created
23customresourcedefinition.apiextensions.k8s.io/networksets.crd.projectcalico.org created
24clusterrole.rbac.authorization.k8s.io/calico-kube-controllers created
25clusterrole.rbac.authorization.k8s.io/calico-node created
26clusterrolebinding.rbac.authorization.k8s.io/calico-kube-controllers created
27clusterrolebinding.rbac.authorization.k8s.io/calico-node created
28daemonset.apps/calico-node created
29deployment.apps/calico-kube-controllers created
30
31$ kubectl get no
32NAME STATUS ROLES AGE VERSION
33master Ready control-plane 87s v1.24.0
The workers can now be connected, so get kubeadm to generate the required command on the master.
1$ kubeadm token create --print-join-command
The run the command generated above on each worker
1$ sudo kubeadmin join ............
The master should now show 3 nodes
1
2$ kubectl get no
3NAME STATUS ROLES AGE VERSION
4master Ready control-plane 120s v1.24.0
5worker-1 Ready <none> 77s v1.24.0
6worker-2 Ready <none> 97s v1.24.0
comments powered by Disqus