🔧 Minikube Remote Access from Local Machine

This guide explains how to access a Minikube Kubernetes cluster running on an AWS EC2 instance from your local machine using a tunnel-based setup.

📘 Prerequisites

• Minikube is installed and running on an AWS EC2 instance.
• You have SSH access to the EC2 server.
• kubectl is installed on your local machine.
• A local file path (e.g., ~/minikube-ai/config) is used to store the remote kubeconfig.

✅ Step-by-Step Guide

🔹 Step 1: Get Minikube kubeconfig from EC2

On your EC2 server (as root or minikube user):

cat ~/.kube/config

Copy the content to a local file on your local machine:

mkdir -p ~/minikube-ai
nano ~/minikube-ai/config
# Paste the contents here

🔹 Step 2: Fix kubeconfig paths (cert/key files)

In the copied config file, update all file paths to absolute paths on your local machine.

You’ll need the following files from EC2:

• /root/.minikube/ca.crt
• /root/.minikube/profiles/ai/client.crt
• /root/.minikube/profiles/ai/client.key

Transfer them using scp:

scp -i key.pem ubuntu@<EC2_PUBLIC_IP>:/root/.minikube/ca.crt ~/minikube-ai/
scp -i key.pem ubuntu@<EC2_PUBLIC_IP>:/root/.minikube/profiles/ai/client.crt ~/minikube-ai/
scp -i key.pem ubuntu@<EC2_PUBLIC_IP>:/root/.minikube/profiles/ai/client.key ~/minikube-ai/

Now update ~/minikube-ai/config to reference the correct local paths:

certificate-authority: /home/<your-user>/minikube-ai/ca.crt
client-certificate: /home/<your-user>/minikube-ai/client.crt
client-key: /home/<your-user>/minikube-ai/client.key

🔹 Step 3: Setup SSH Tunnel

Create a local port-forward tunnel to Minikube’s Kubernetes API Server:

ssh -i key.pem -L 8443:192.168.49.2:8443 ubuntu@<EC2_PUBLIC_IP>

• 192.168.49.2 is Minikube’s internal IP.
• Keep this terminal running. Or run it with autossh or as a background job.

🔹 Step 4: Update API Server URL in kubeconfig

Edit your ~/minikube-ai/config:

server: https://localhost:8443

This tells kubectl to connect to the local port from your SSH tunnel.

🔹 Step 5: Set kubeconfig and default namespace

You can either pass --kubeconfig every time or set it globally:

export KUBECONFIG=~/minikube-ai/config

👉 Add this to your ~/.bashrc or ~/.zshrc:

echo 'export KUBECONFIG=~/minikube-ai/config' >> ~/.bashrc
source ~/.bashrc

Set the default namespace (ai-assistant in this case):

kubectl config set-context ai --namespace=ai-assistant

✅ Final Test

Once the SSH tunnel is active and everything is configured:

kubectl get pods

You should see pods running in the ai-assistant namespace.

🔁 Troubleshooting

🔐 Security Notes

• Only open port 22 on EC2 (no need to expose Kubernetes API publicly).
• Always use an SSH key with limited access.
• Optionally, restrict SSH to specific IPs via AWS Security Groups.