Document for this devops test

Topics

• Run the test
• Provision Kubernetes Cluster
• Creating Helm Charts
• Create Single Executator File

Run the test

First you must clone this repository with the command below and change directory to it:

git clone https://github.com/PooriaShahi/devops-test.git
cd devops-test

Then make sure you have pip3 installed on your system. you can check that with the command below:

pip3 -V

If it's not installed on your system then do the below: Python installation

python -m pip3 install --upgrade pip

Ubuntu installation

sudo apt install python3-pip

MacOS installation

brew install pip3

Then make sure that you have a server with root user and you have a ssh-key based authentication with it.

For your domain, please change value of host in wart/values.yaml file.

Then run the command below:

bash automate.sh

Enter you server's IP and go make some snack and some coffee until magic completely happen ;)

When magic ends, you can see 2 different url based on your domain that defined in the wart/values.yaml file. Copy the urls and see the results.

Provision Kubernetes Cluster

For this challenge i use kubespray as recommended and provision a single node (CP+Worker) k8s cluster. I using calico CNI plugin for overlay k8s network. First creating my variables with the command below:

cd kubespray
cp -rfp ./kubespray/inventory/sample ./kubespray/inventory/mycluster

add inventory.ini add this configuration here

#[kube_control_plane]
#node1 ansible_host=nl-rv.maxtld.com ansible_port=3679 ansible_user=mohammad ip=10.254.11.134 etcd_member_name=etcd1

#[etcd:children]
#kube_control_plane

#[kube_node]
#node1

#[all:vars]
#ansible_ssh_user=mohammad
#ansible_become=true
#ansible_become_password='add your password'

# This inventory file is for a single-node Kubernetes cluster with etcd running on the same node as the control plane

[all]
server1 ansible_host=nl-rv.maxtld.com ansible_port=3679 ansible_user=mohammad ip=10.254.11.134 etcd_member_name=etcd1

[kube_control_plane]
server1

[etcd]
server1

[kube_node]
server1

[k8s-cluster:children]
kube_control_plane
kube_node

[all:vars]
ansible_ssh_user=mohammad
ansible_become=true
k8s_version=v1.21.0
network_plugin=calico  # You can change the network plugin to another one if needed

after that add your ssh-key

ssh-copy-id -p 3679 [email protected]

then when in kubespray dir just run:

ansible-playbook -i inventory/mycluster/inventory.ini  cluster.yml -b -v --private-key=~/.ssh/id_rsa -K -u mohammad

after a couple of mins your cluster is ready

notice in ansible.cfg we set that all our command run with sudo

we get admin.conf for kubectl local-machine but for intraction we tunnel traffic between node and local-machine

ssh -f -N -L 6443:10.254.11.134:6443 -p 3679 [email protected]

for using kubeconfig:

echo "export KUBECONFIG=/home/mohammad/.kube/admin.conf" >> ~/.bashrc
source ~/.bashrc

Creating Helm Charts

for create helm chart

helm create wordpress-mysql-pma

after that add our configuration

In this chart for every objects, manifest written and we deploy the chart with the command below:

helm install wordpress-mysql-pma ./wordpress-mysql-pma

for down the deployment

helm uninstall wordpress-mysql-pma ./wordpress-mysql-pma

Create Single Executator File

we have pre-install job for update and upgrade node jobs/pre-install.ini. All this will be done with a automate.sh file which is a bash script file for implementing all together. for destroy cluster use destroy.sh