In this tutorial you are going to learn How to setup K3S Cluster in EC2 and Join worker nodes and deploy a simple JavaScript Game.
First we need to create EC2 Instances.If you don't know how to create an EC2 Instance you can check this article.
Before You Start
You need
- 3 EC2 instances (1 master, 2 workers)
- Configured with permissions for the necessary ports (e.g. SSH, K3s API, NodePort).
- AWS CLI configured for your account.
Hardware Requirements:
Hardware requirements scale based on the size of your deployments. Minimum recommendations are outlined here.
1. Minimum Hardware Requirements for Master Node:
The master node in a Kubernetes cluster manages the control plane, handling tasks like scheduling, maintaining cluster state, and managing the API server. As such, it requires adequate resources to perform these tasks efficiently.
- CPU: 2 vCPUs (minimum)
- RAM: 2 GB (minimum); 4 GB or more recommended for production environments
- Storage: 10 GB (minimum)
2. Minimum Hardware Requirements for Worker Nodes:
Worker nodes in a Kubernetes cluster host the application workloads, running containers that are orchestrated by the master node. The requirements for worker nodes depend on the workloads they will run, but the following are the general minimum recommendations:
- CPU: 1 vCPU (minimum); more depending on workload demands
- RAM: 1 GB (minimum); 2 GB or more recommended for moderate workloads
- Storage: 10 GB (minimum)
- Network: High-speed and low-latency network connectivity, especially for communication with the master node and other worker nodes
Example Configuration for AWS EC2:
- Master Node: t3.medium (2 vCPUs, 4 GB RAM) for a small cluster
- Worker Nodes: t3.small (2 vCPUs, 2 GB RAM) or t3.medium for moderate workloads
We can see we have 3 Instances running
Inbound Rules need to allow for K3s Nodes
Protocol | Port | Source | Destination | Description |
---|---|---|---|---|
TCP | 2379-2380 | Servers | Servers | Required only for HA with embedded etcd |
TCP | 6443 | Agents | Servers | K3s supervisor and Kubernetes API Server |
UDP | 8472 | All nodes | All nodes | Required only for Flannel VXLAN |
TCP | 10250 | All nodes | All nodes | Kubelet metrics |
UDP | 51820 | All nodes | All nodes | Required only for Flannel Wireguard with IPv4 |
UDP | 51821 | All nodes | All nodes | Required only for Flannel Wireguard with IPv6 |
TCP | 5001 | All nodes | All nodes | Required only for embedded distributed registry (Spegel) |
TCP | 6443 | All nodes | All nodes | Required only for embedded distributed registry (Spegel) |
- The K3s server needs port 6443 to be accessible by all nodes.
- Typically, all outbound traffic is allowed.
Step 1: Install K3S on Master Node
In this tutorial I am using Ubuntu 22.04. If you are using any old version of Ubuntu you may face some issues with iptables.
Check more here:
Now
- 1.0 SSH your Master node (i am using web console)
Run Following Commands on terminal.
- 1.3. Now Install k3s on master node.
Command: curl -sfL https://get.k3s.io | sh -
- 1.4 Check Status
systemctl status k3s
Now Get the node token (needed to join worker nodes):
You need this token to securely join worker nodes to the master node (also known as the control plane).
For Example
- Authentication
- Secure Communication
- Cluster Management
- Simplified Node Addition
Command
sudo cat /var/lib/rancher/k3s/server/node-token
Step 2: Prepare Worker Node
Same Above we need to create 2 EC2 instances for worker nodes. For this tutorial we have two t3.medium EC2 machine. Now SSH both worker nodes and perform the below tasks:
Install Docker in the both worker machine.
Join the machine in master node using the token.
- 2.0 Change the host name to Workwer-1 & Worker-2
- 2.1 Install Docker
Commands:
sudo su
apt update
apt install docker.io -y
2.2 Check Status
systemctl status docker
Worker-2
Step 3: Join Worker Nodes to the Cluster (Master Node)
First get the token form master node. We already collect the token earlier this tutorial. Now to join the nodes in master nodes run this command from worker nodes.
Command
curl -sfL https://get.k3s.io | K3S_URL=https://<MasterNodePublicIP>:6443 K3S_TOKEN=<NodeToken> sh -
Here change the
<MasterNodePublicIP>
with your EC2 master node Public IP &<NodeToken>
with your token from master node.
Worker-1
Now Check the nodes list from master node
Command: kubectl get nodes
Step 4: Now Deploy a JS Game on k3s:
To run an application on K3s, we need to create Kubernetes deployment and service files.
First Create a Namespace
Kubernetes namespace is like a separate section within a Kubernetes cluster, where you can keep your resources organized and isolated from others. Think of it as different folders on your computer. Each namespace (or folder) can have its own set of applications, settings, and permissions, and they won’t interfere with each other.
Command: kubectl create namespace js-game
Now Create Kubernetes Deployment File:
Command: nano js-game-app.yml
Now Copy Paste the code into js-game-app.yml
in the Image line you can update your dockerhub or any image link.
Applying the Manifest
Save this configuration to a file named js-game-app.yml and apply it using the following command: kubectl apply -f js-game-app.yml -n js-game
`
You will see this if everything is work correctly....
Verify the Deployment
To check if the deployment, service, and ingress are created successfully, you can use the following commands:
kubectl get deployments -n js-game
kubectl get services -n js-game
kubectl get ingress -n js-game
Now check the pods running....
Command: kubectl get pods -n js-game
Now open the browser and visit with your Master Node Public IP
Ref:
- Docker Image: internaldev/jsgame:latest
- YAML File: https://github.com/mimqupspy/k3s-yaml
Top comments (0)