kubetwo 0.1.0

Simple CLI tool to create Kubernetes cluster on AWS EC2.

kubetwo

kubetwo is a simple CLI tool to create a Kubernetes cluster on AWS EC2 with just one command.

kubetwo will interact with Terraform and Ansible (Kubespray) for provisioning.

Requirements

To use kubetwo, following tools are necessary.

• Python (3.6 and newer)
• Terraform (1.0.0 and newer)

Quick Start

1. Prerequisites

First, please install kubetwo package using pip or pip3 depending on your environment.

pip install kubetwo

For kubetwo to interact with Terraform and create AWS resources, it's necessary to set environment variables about AWS credentials.

export AWS_ACCESS_KEY_ID="your_anaccesskey"
export AWS_SECRET_ACCESS_KEY="your_asecretkey"
export AWS_DEFAULT_REGION="your_region"

Then, you need to create ssh key to access the EC2. If you've already used ssh key, you can use it.

ssh-keygen -t rsa -b 4096 -f ~/.ssh/kubetwo_id_rsa

2.1 Spin up Kubernetes cluster (kubetwo apply)

By running the following kubetwo apply command, kubetwo will spin-up AWS EC2 with Terraform and set Kubernetes cluster on them with Ansible (Kubespray v2.16.0).

kubetwo apply \
--cluster_name="kubetwo-cluster" \
--ssh_public_key="~/.ssh/kubetwo_id_rsa.pub" \
--ssh_private_key="~/.ssh/kubetwo_id_rsa" \
--ami="ami-0df99b3a8349462c6" \
--control_plane=1 \
--worker_node=2 \
--open_ports="[6443, 30080]" \
--availability_zone="ap-northeast-1a" \
--deploy_sample

By using --approve option, you can automate approval. To know more details about parameters, you can refer to the following section.

Note: Port 6443 will be used for kube-apiserver and port 30080 for sample nginx service.

2.2 Spin up Kubernetes cluster (kubetwo init & create)

Instead of using kubetwo apply command, you can use kubetwo init and kubetwo create commands separately.

First, kubetwo init command will create the workspace in the current directory and render Terraform manifests and inventory file for Ansible on it. Also, it will download the Kubespray archive from GitHub.

kubetwo init \
--cluster_name="kubetwo-cluster" \
--ssh_public_key="~/.ssh/kubetwo_id_rsa.pub" \
--ami="ami-0df99b3a8349462c6" \
--control_plane=1 \
--worker_node=2 \
--open_ports="[6443, 30080]" \
--availability_zone="ap-northeast-1a"

Then, kubetwo create command will spin up EC2 and Kubernetes cluster.

kubetwo create \
--cluster_name="kubetwo-cluster" \
--ssh_private_key="~/.ssh/kubetwo_id_rsa" \
--deploy_sample

3. Access to the Kubernetes cluster

After the Kubernetes cluster is created, you can see admin.conf in the workspace. Setting KUBECONFIG will allow you to access the Kubernetes cluster.

export KUBECONFIG=$(pwd)/kubetwo_cluster/admin.conf

Let's check the Kubernetes cluster from your local machine. (If you don't have kubectl, please install it)

kubectl get nodes

Also, you can check nginx sample if you add --deploy_sample option. URL will be shown when kubetwo create is completed.

curl http://xx.xx.xx.xx:30080

4. Clean up created resources

You can clean up created resources with kubetwo delete command. It will also delete workspace for kubetwo on your machine.

kubetwo delete \
--cluster_name="kubetwo-cluster"

Parameters

kubetwo apply

Name	Type	Default	Description
cluster_name	string	-	Kubernetes cluster name
ssh_public_key	string	-	Path of public key for SSH
ssh_private_key	string	-	Path of private key for SSH
ami	string	ami-0c3fd0f5d33134a76	Amazon Machine Image of EC2 instance
instance_type	string	t3.medium	Instance type of EC2 instance
availability_zone	string	us-west-1a	Availability zone of EC2 instance
control_plane	int	1	Number of control planes
worker_node	int	1	Number of worker nodes
open_ports	array[string]	[6443, 30080]	Ports to open for global (If you use -1, all ports will be open)
deploy_sample	boolean	false	If true, kubetwo will deploy nginx sample in cluster
approve	boolean	false	If true, kubetwo won't prompt you to approve for execution

kubetwo init

Name	Type	Default	Description
cluster_name	string	-	Kubernetes cluster name
ssh_public_key	string	-	Path of public key for SSH
ami	string	ami-0c3fd0f5d33134a76	Amazon Machine Image of EC2 instance
instance_type	string	t3.medium	Instance type of EC2 instance
availability_zone	string	us-west-1a	Availability zone of EC2 instance
control_plane	int	1	Number of control planes
worker_node	int	1	Number of worker nodes
open_ports	array[string]	[6443, 30080]	Ports to open for global (If you use -1, all ports will be open)
approve	boolean	false	If true, kubetwo won't prompt you to approve for execution

kubetwo create

Name	Type	Default	Description
cluster_name	string	-	Kubernetes cluster name
ssh_private_key	string	-	Path of private key for SSH
deploy_sample	boolean	false	If true, kubetwo will deploy nginx sample in cluster
approve	boolean	false	If true, kubetwo won't prompt you to approve for execution

kubetwo delete

Name	Type	Default	Description
cluster_name	string	-	Kubernetes cluster name
approve	boolean	false	If true, kubetwo won't prompt you to approve for execution