Configure K8S Multi-Node Cluster over AWS using Ansible
Lets first introduce about the file structure that
“How I am going to configure it ?”
So, I have made the four roles :-
1. For launching the instances
2. Configuring the Common Setup which need on Both Nodes i.e master and slave.
3. Configuring the master node (like initializing master , configuring flannel etc . . . .)
4. Configuring the Slave node ( like joining with cluster . . . .)
If You Don’t Know About role or How to convert your playbook to role ?
First I would recommend that you should must read the given blog first then proceed . . .
How we can convert our ansible PlayBook to Role ? 🤔
We Will Also Going To Share that on Ansible Galaxy !! 🤔
Lets Start !
Step 1: First Launch the Instance for the master and slave (according to your requirement).
In my case I am going to launch 1 instance for the master and 2 or 3 for the slave. Link of the role given below :
Ansible Galaxy
Jump start your automation project with great content from the Ansible community
Shashwatsingh22/instance_ec2
A brief description of the role goes here. Any pre-requisites that may not be covered by Ansible itself or the role…
You can change the variable according your requirement .
Master →
Slave →
Run the Play →
Step 2: Understand the Second Role Which will configure the both instances master as well as slave node .
Shashwatsingh22/k8s_cluster
Ansible Collection — . Contribute to Shashwatsingh22/k8s_cluster development by creating an account on GitHub.
Ansible Galaxy
Jump start your automation project with great content from the Ansible community
1> First Configure , Install and Start the service Docker.
2> Configuring the yum for the installation of the required software’s like kubectl , kubelet , kubeadm …
3> Disabling the Security of the SElinux for the smooth configuration k8s cluster .
4> Installing the Important Software’s like kubeadm, kubelet ,kubectl and Starting the kubelet Service.
5> Pulling the images which will manage our cluster.
6> Changing the Driver of the Docker and restarting the docker service.
6> Installing the iproute-tc
Step 3: Now the third role come in play which will configure the master.
1> Setting bridge-nf-call-iptables = 1
2> Initializing Master
3> Creating the .kube directory ,
Copying the /etc/k8s/admin.conf$HOME/.kube/config &&
Changing the owner permission of $HOME/.kube/config
4> Genrating token and storing it . When it need to join the slave to cluster.
5> Adding the Flannel Overlay NetWork for the Communication.
Step 3: Now the Fourth role come in play which will configure the slave.
1> Setting bridge-nf-call-iptables & ip6tables to 1.
2> Retrieving the token and joining the slave to the Cluster
Now ! Here We have complete the SetUp Of the K8s Cluster .
→Lets Go And Interact with master and run the command to ensure that
“How many nodes are connected the master ?”
$ kubectl get nodes
Hope You Enjoy this Automation ! !
Thank You !!
Shashwat Singh
Cyber Security Enthusiasm || Docker || RHEL8 || Working On Web Development || CTF Player || Machine Learning || Flutter Developer || Ansible || K8s
Follow
1
1
More from Shashwat Singh
Follow
Cyber Security Enthusiasm || Docker || RHEL8 || Working On Web Development || CTF Player || Machine Learning || Flutter Developer || Ansible || K8s
Create An Ansible Playbook to Retrieve the Container IP and update the Inventory File & Configure the Webserver.
Lets First Build An Docker Image In Such A way so that we can do SSH to it . . .
We can do either Manually or by the help of DockerFile ;
Configure The LoadBalancer On EC2 Instance !!
Lets Start😎 !!
Lets Write an Dynamic Playbook In Such a way we can lunch the Instance as Per Our Need.
→ Launch the 3 Instances for the Web-Server -
- hosts: localhost
become: Falsevars_prompt:
- name: noOS
private: no
prompt: "How Many Instances Do You Wants Lunch ?"vars_files:
- awsAccess.ymltasks:
- name: Launch instance for webserver
ec2:
count: "{{ noOS }}"
group_id: "sg-0d4c5f0d452e1b620"
image: "ami-08e0ca9924195beba"
aws_access_key: "{{ usern }}"
aws_secret_key: "{{ pass }}"
instance_type: "t2.micro" …
