Skip to content

Deployment & Hosting

The kube-q web terminal is a Docker container. You need somewhere to run it so users can reach it from a browser.

Hosting options

Platform Free tier WebSockets Always-on Docker Verdict
Oracle Cloud 4 CPU / 24 GB RAM — forever ✓ ✓ ✓ Most generous, manual setup
Fly.io 3 shared VMs ✓ ✓ ✓ Best managed option
Railway $5 credit/mo ✓ ✓ ✓ Easiest setup
Render Yes ✓ ✗ (sleeps after 15 min) ✓ Bad for a terminal
Koyeb Yes ✓ ✓ ✓ Good alternative

Oracle Cloud has the most generous free tier of any cloud provider — it never expires and never charges you. The only trade-off is manual setup.

What you get — free forever

  • 4 ARM CPU cores + 24 GB RAM (Ampere A1) — use all 4 on one VM
  • 2 AMD VMs (1 CPU, 1 GB RAM each)
  • 200 GB storage
  • No time limit, no credit card expiry

A single ARM VM with 2 cores and 12 GB RAM is more than enough for the kube-q web terminal.

Setup

1. Create a free account

Go to cloud.oracle.com → sign up → select your home region (pick one close to you — you cannot change it later).

2. Create a free VM

  • Compute → Instances → Create Instance
  • Shape: Ampere A1 (ARM) → 2 OCPUs, 12 GB RAM
  • Image: Ubuntu 22.04
  • Add your SSH public key

3. Install Docker

ssh ubuntu@<your-vm-ip>

sudo apt update && sudo apt install -y docker.io
sudo usermod -aG docker ubuntu
# log out and back in for the group to take effect

4. Open port 3000

In Oracle Console → Networking → VCN → Security List → add an ingress rule for TCP port 3000.

Also allow it in the VM's local firewall:

sudo iptables -I INPUT -p tcp --dport 3000 -j ACCEPT

5. Run the container

docker run -d \
  --name kube-q \
  --restart always \
  -p 3000:3000 \
  -e KUBE_Q_URL=https://your-backend.com \
  -e KUBE_Q_API_KEY=your-key \
  ghcr.io/mskazemi/kube_q:latest

Open http://<your-vm-ip>:3000 — done.

Install nginx and Certbot for a proper domain + TLS:

sudo apt install -y nginx certbot python3-certbot-nginx

# Add your domain's A record pointing to the VM IP first, then:
sudo certbot --nginx -d kube-q.example.com

Nginx config (/etc/nginx/sites-available/kube-q):

server {
    listen 443 ssl;
    server_name kube-q.example.com;

    location / {
        proxy_pass http://localhost:3000;
        proxy_http_version 1.1;
        proxy_set_header Upgrade $http_upgrade;
        proxy_set_header Connection "upgrade";
        proxy_set_header Host $host;
    }
}

Note

The Upgrade and Connection headers are required for WebSocket (PTY) connections to work through nginx.

Fly.io (easiest managed option)

Fly.io deploys directly from your Dockerfile with one command and provides automatic HTTPS.

# Install flyctl
curl -L https://fly.io/install.sh | sh

# From the kube_q project root
fly launch        # detects Dockerfile automatically
fly secrets set KUBE_Q_URL=https://your-backend.com
fly secrets set KUBE_Q_API_KEY=your-key
fly deploy

Your app will be live at https://kube-q.fly.dev (or a name you choose). Free tier includes 3 shared VMs with automatic HTTPS and built-in WebSocket support.

Railway

Connect your GitHub repo and Railway auto-deploys on every push to main.

  1. Go to railway.app → New Project → Deploy from GitHub repo
  2. Select MSKazemi/kube_q
  3. Add environment variables: KUBE_Q_URL, KUBE_Q_API_KEY
  4. Railway detects the Dockerfile and builds automatically

Free tier includes $5 credit/month — enough for light usage.

Oracle vs Fly.io comparison

Oracle Cloud Fly.io
Setup effort ~30 min manual 5 min, one command
Resources 4 CPU / 24 GB RAM Shared, limited
Cost Free forever Free tier, then pay-as-you-go
HTTPS Manual (nginx + Certbot) Automatic
Auto-deploy on push Manual (you set up CI) Built-in

Oracle is better long-term value (much more RAM/CPU, truly free forever).
Fly.io is faster to get running with less ops work.