kube-vagrant

In this Project We'll Deploy Kubernetes Cluster and then Monitor it using Prometheus and Grafana

Step One: Deploy Kubernetes multi master ha cluster with Vagrant and virtualbox.

Step Two: Using Ceph-Rook as Storage Class.

Step Three: At the End Deploy Mariadb HA with Galera Cluster and Monitor Deploymet by using prometheus, thanos and grafana.

All steps will be automated

Requirements

• Host with 24 core CPU and 48 Gi RAM and any OS nested virtualization will be enabled.
• 3 Instance(Ubuntu 20.04) with 6 core CPU and 12 Gi RAM. Every 3 kubernetes nodes are also ControlPlane(master) and Worker.
• Download and install Vagrant in host from https://www.vagrantup.com/downloads .
• Download and install VirtualBox from https://www.virtualbox.org/wiki/Downloads .

Attention : we use https://shecan.ir/ DNS in this project for lifting of sanctions

Ansible

Usage

Install Ansible according to Ansible installation guide then run the following steps:

Before Deploy

You need to install vagrant-disksize plugin for extra disk cause ceph rook need second hard disk drive in raw (without format) for install please run in host:

vagrant plugin install vagrant-disksize

You can edit Vagrant file variables in vagrant/config.rb before deploy

Quick Start

To deploy the cluster you can use :

Vagrant

Usage

# Install dependencies from ``requirements.txt``
sudo pip install -r requirements.txt
vagrant up

If a vagrant run failed and you've made some changes to fix the issue causing the fail, here is how you would re-run ansible:

ansible-playbook -vvv -i .vagrant/provisioners/ansible/inventory/vagrant_ansible_inventory --become --become-user=root cluster.yml

Access to cluster nodes and manage

vagrant ssh k8s-1
sudo cp /etc/kubernetes/admin.conf .
sudo chown vagrant:vagrant admin.conf
export KUBECONFIG=admin.conf

kubectl get nodes -o wide

Check cluster nodes state

kubectl get nodes -o wide

About the solutions use in this challenge

for deploying we use [kubespry] https://github.com/kubernetes-sigs/kubespray.git in this project we have changed features and added new features :

• Use shecan dns by adding role in roles/ for lifting of sanctions
• Change in Vagrantfile for nodes extra disk
• Add rook_ceph role in cluster.yaml and create rook role in roles/kubernetes-app/ for installing rook as kubernetes storage class
• Add galera_cluster role in cluster.yaml and create galera_cluster role in roles/kubernetes-app/ for installing galera cluster database on storage class
• Add monitoring role in cluster.yaml and create monitoring role in roles/kubernetes-app/ for installing prometheus and grafana and monitor mysql metrics in dashboard with helm