I built this project to create my own lab for Kuberntes cluster on AWS cloud using EC2 instances. I found Terraform is the best tool to create my K8S lab fastly with one command ๐.
- EC2
- One Master Node
- Two Worker Node (can be increased)
- VPC
- Public Subnet
- Internet Gateway
- Route Table
- Security Group
- S3 Bucket
The goal is to build a K8S cluster with one master node and two worker nodes.
- First, the master node will boot up and will start installing kubeadm, kubelet, kubectl, and docker. Then will run
kubeadm init
to initial the k8s cluster.
Here the challenge becomes how we can get the join command that showed after init the cluster and send it to the workers node for joining the worker node into the cluster ๐ค?
To solve this problem I use s3 bucket. First I extracted the join command and saved it into a file, and then pushed it to the s3 object. Now we finish from the master node and is ready.
- Second, the workers node will boot up and will start installing kubeadm, kubelet, kubectl, and docker. Then will fetch the joined command from s3 bucket and execute to join the worker node into the cluster.
- By default there are two workers on the cluster, to increase it go to
variables.tf
file and looking for number_of_worker variable, you can increase the default number.
1- Make sure you have the Terraform tools installed on your machine.
2- Add your Access key, Secret key, and Key Pair name on variables.tf
file.
3- Make sure your IAM user has the right permission to create EC2, VPC, S3, Route Table, Security Group, and Internet Gateway.
After doing the requirements, you are ready now, start cloning the repo to your machine:
git clone https://github.com/Ahmad-Faqehi/Terraform-Bulding-K8S.git
cd Terraform-Bulding-K8S/
Now execute terraform commands:
terraform init
terraform plan #to show what going to build
terraform apply
- You can access your cluster by accessing the master node through ssh, you can get the public IP of the master node from terraform outputs. Below is an example of the ssh command:
ssh -i <Your_Key_Piar> ubuntu@<MasterNode_Public_IP>
- Another way to access the cluster is by downloading the
admin.conf
file from the master node to your machine, find below the way to download it and access the cluster remotely.
scp -i <Your_Key_Piar> ubuntu@<MasterNode_Public_IP>:/tmp/admin.conf .
This will download the Kubernetes config file on your machine. Before using this config file, you have to replace the private IP with the public IP of the master node. Then you can use the following command to start accessing the cluster.
kubectl --kubeconfig ./admin.conf get nodes
To destroy all the resources that were created after applying the script, just run the following command:
terraform destroy