Cluster Kubernetes with Vagrant

---

---

Structure and technologies used

What purpose?

This cluster has as main objective the learning of technologies, initially at the IAC level and later for DevOps process studies.

Technologies used:

• Vagrant
• Docker
• ShelScript
• Kubernetes
• VirtualBox

Basic structure:

4x VM's, each containing: 2GB of memory, 2vCPU, 40GB of dynamically allocated Disk, network interface in bridge mode and for operating system, Ubuntu Bionic Beaver 18.04 LTS.

Basic topology:

---

---

How to use

What hardware is needed on the computer?

12GB of memory, processor with 8 cores (physical or virtual), 20GB of available disk space and a network adapter with internet connection.

Additional remarks *

In case you don't have the specified amount of hardware, you can remove the configuration block in the Vagrantfile for worker03. Do not modify the CPU or Memory values ​​of the cluster nodes to less than what is already configured!

What do I need to have installed on my computer?

• Vagrant
• VirtualBox

Files to be modified:

Modify the value of "advertiseAddress" in kubeadm-config.yaml with your network information, as follows:

...
  advertiseAddress: "your-ip-address-for-k8smain01"
...

Modify the value of "bridge" and "ip" in Vagrantfile with your network information, as follows:

    ...
    k8smain01.vm.network "public_network",
     bridge: "your-network-interface-name",
     ip: "your-ip-address-for-k8smain01"
    ...
    k8sworker01.vm.network "public_network",
     bridge: "your-network-interface-name",
     ip: "your-ip-address-for-k8sworker01"
    ...
    k8sworker02.vm.network "public_network",
     bridge: "your-network-interface-name",
     ip: "your-ip-address-for-k8sworker02"
    ...
    k8sworker03.vm.network "public_network",
     bridge: "your-network-interface-name",
     ip: "your-ip-address-for-k8sworker03"
    ...

Additional remarks *

To get the interface name information, use the commands:

• GNU/Linux: ip a | grep mtu
• macOs: ifconfig | grep mtu
• Windows: ipconfig /all | find /i "Description"

Use an interface from your routed local network (which has internet access) to define the IP address of the VM's!

Files that will be modified automatically:

The file kubeadm-join.sh will be modified automatically, no changes are needed in this file, it will receive data as follows:

kubeadm join <ip-of-k8smain01>:6443 --token <token> --discovery-token-ca-cert-hash sha256:<discovery-token>

How to run the project on my computer?

• Download the project in the github root repository cluster-k8s.
• Inside the directory, open the terminal and run the command vagrant up.
• If you have any problems, use vagrant destroy to delete the VM's.
• The vagrant halt command will just stop the VM's.

---

---

Validation commands

List of commands and their outputs:

Using Vagrant's SSH utility (vagrant ssh) access VM k8smain01 (vagrant ssh k8smain01), and use the following list of commands to validate cluster operation:

Command	Applicability
kubectl get nodes -o wide	Node state information
kubectl get namespaces	Information about existing namespaces
kubectl get pods -n kube-system -o wide	Make sure the kube-proxy pod exists and the state is Running
kubectl describe pod yourPodName -n kube-system	Specific pod events
kubectl logs yourPodName -n kube-system	Specific pod logs
kubectl describe node nodeName	information about the state of a specific node

---

---

References

Documentation used to create this project:

• https://e-tinet.com/linux/vagrant/
• https://medium.com/@fidelissauro/kubernetes-criando-um-cluster-simples-com-o-kubeadm-e50a9eb4f4a6
• https://www.weave.works/docs/net/latest/kubernetes/kube-addon/#pod-network
• https://www.youtube.com/watch?v=CNZmKjklfeg

---

---

end of README.md