Kubernetes Installation on CentOS 8

We have covered the basic overview of Kubernetes. Now we will cover its installation on CentOS 8. Let us first cover what CentOS 8 is.

CentOS 8 is a Linux distribution which gives us a free and open source community and computing platform, released on 24 Sep 2019. It is compatible with Red Hat Enterprise Linux and with its upstream source. From version 8 it supports x86-64, ARM64, and POWER8 architectures.

The different ports which are used for access and communication are:

Prepare Firewall and HostName

Prerequisites for Master and Worker Nodes

1. The minimum requirements for the Kubernetes server and for master and worker nodes are 2GB and CPUs. Kubernetes cannot be installed until these minimum requirements are met.

2. Before installing anything you need to update the servers using:

dnf -y upgrade

3. Disable SELinux enforcement.

1
2
3
setenforce 0
sed - i--follow - symlinks 's/SELINUX=enforcing/SELINUX=disabled/g'
etc / sysconfig / selinux

4. The IP masquerade at the firewall has to be enabled using:

1
2
firewall - cmd--add - masquerade--permanent
firewall - cmd--reload

5. To traverse iptables rules we need to set bridged packets.

1
2
3
4
cat < /etc/sysctl.d / k8s.conf
net.bridge.bridge - nf - call - ip6tables = 1
net.bridge.bridge - nf - call - iptables = 1
EOF

6. Then we need to load new rules.
   sysctl --system

7. To increase performance, disable all the memory swaps.
   swapoff -a

After the above mentioned steps we can now install Docker.

Installing Docker on Master and Worker Nodes

1. To install Docker add the repository for the Docker installation package.

1
dnf config - manager--add - repo = https: //download.docker.com/linux/centos/docker-ce.repo

2. Before installing Docker install container.io.

1
dnf install https: //download.docker.com/linux/centos/7/x86_64/stable/Packages/containerd.io-1.2.6-3.3.el7.x86_64.rpm

3. Install Docker from the repository.
   dnf install docker-ce --nobest -y

4. Now we need to start Docker service.
   systemctl start docker

5. To make Docker auto start when server is started:
   systemctl enable docker

6. To use systemd cgroup driver to make changes to Docker:

1
2
3
4
echo '{
"exec-opts": ["native.cgroupdriver=systemd"]
}
' > /etc/docker/daemon.json

7. To apply the changes restart Docker.
   systemctl restart docker

8. Check the version of Docker using:
   docker version

Installing Kubernetes on Master and Worker Nodes

After installing Docker now we will install Kubernetes.

1. Create the file to add the Kubernetes repository to the package.

1
2
3
4
5
6
7
8
9
cat < /etc/yum.repos.d / kubernetes.repo[kubernetes]
name = Kubernetes
baseurl = https: //packages.cloud.google.com/yum/repos/kubernetes-el7-\$basearch
  enabled = 1
gpgcheck = 1
repo_gpgcheck = 1
gpgkey = https: //packages.cloud.google.com/yum/doc/yum-key.gpg https://packages.cloud.google.com/yum/doc/rpm-package-key.gpg
  exclude = kubelet kubeadm kubectl
EOF

2. Update the repository info.
   dnf upgrade -y

3. To install the necessary components for Kubernetes:
   dnf install -y kubelet kubeadm kubectl --disableexcludes=kubernetes

4. Enable Kubernetes services to run at startup and start its services.

1
2
systemctl enable kubelet
systemctl start kubelet

Configuring Kubernetes on Master Node

After its installation we will configure it to form a cluster.

1. Configure kubeadm.
   kubeadm config images pull

2. Open the ports which are used by Kubernetes.
   firewall-cmd --zone=public --permanent --add-port={6443,2379,2380,10250,10251,10252}/tcp

3. Replace the worker-IP-address to allow Docker access from another node.
   firewall-cmd -zon=public --permanent -add-rich-rule 'rule family=ipv4 source address=worker-IP-address/32 accept'

4. Mark the changes made and reload.
   firewall-cmd --reload

5. Install the plugin Container Network Interface (CNI) for Kubernetes.
   kubeadm init --pod-network-cidr 192.168.0.0/1

You will then get an output like shown below. Please make a note of the token as it is required to join worker nodes to the cluster.

Output:

6. Now make the configuration files and directory.

1
2
3
4
mkdir - p $HOME / .kube
cp - i / etc / kubernetes / admin.conf $HOME / .kube / config
chown $(id - u): $(id - g) $HOME / .kube / config
kubectl apply - f https: //docs.projectcalico.org/manifests/calico.yaml

7. Check that the master node is enabled and is running.
   kubectl get nodes

After successful execution, we will get a node with ready status.

1
2
NAME STATUS ROLES AGE VERSION
master Ready master 91 s v1 .18 .0

Configuring Kubernetes on Worker Node

For Kubernetes installation we should have at least one worker node because it runs the containerized applications. In this section we will configure only one worker node.

1. Open all the ports used by Kubernetes.
   firewall-cmd --zone=public --permanent --add-port={10250,30000-32767}/tcp

2. Mark the changes made and reload.
   firewall-cmd --reload

3. Now we need to join a cluster using the token we previously noted.

1
2
kubeadm join 94.237 .41 .193: 6443--token 4 xrp9o.v345aic7zc1bj8ba\
  --discovery - token - ca - cert - hash sha256: b2e459930f030787654489ba7ccbc701c29b3b60e0aa4998706fe0052de8794c

4. Check if the worker node has joined or not.
   Go to the Master node and issue the below command.

kubectl get nodes

1
2
3
NAME STATUS ROLES AGE VERSION
master Ready master 10 m v1 .18 .0
worker Ready 28 s v1 .18 .0

If it all went well, you will see two nodes with status Ready. If It does not happen, wait for a while and run the command again.

Now that it’s all done, you are all ready to work on Kubernetes installation having two nodes.