VPS Deployment

Running on a VPS

If you want to run OpenClaw on a VPS, you don't need a large machine. A Hetzner CX23 is plenty.

I'd strongly recommend Tailscale on both your local machine and the VPS. Install it on the VPS with the --ssh=true flag and you can log in over Tailscale directly. Then block port 22 entirely in the Hetzner firewall.

I block all inbound traffic and access everything over Tailscale. Simple setup, and one less thing to worry about.

Hardware requirements

Network security with Tailscale

Step 1: Install Tailscale on local machine

# On your local machine (Mac/Linux/Windows)
# Visit: https://tailscale.com/download
# Install and authenticate

Step 2: Install Tailscale on VPS

# SSH into VPS normally first
ssh user@your-vps-ip

# Install Tailscale
curl -fsSL https://tailscale.com/install.sh | sh
sudo tailscale up --ssh=true

The --ssh=true flag enables SSH over Tailscale.

Step 3: Get VPS Tailscale IP

# On VPS, get your Tailscale IP
tailscale ip -4
# Example output: 100.64.1.2

Step 4: TEST SSH via Tailscale

# Use the Tailscale IP, NOT your public IP
ssh user@100.64.1.2

Step 5: Block public access (only after verification)

# 1. Create new firewall rule
# 2. Block: All inbound traffic on port 22
# 3. Leave Tailscale traffic unrestricted (100.64.0.0/10)

Step 6: Verify lockout prevention

# This should work (Tailscale)
ssh user@100.64.1.2

# This should fail/timeout (public IP blocked)
ssh user@your-public-vps-ip

OpenClaw installation

After installing OpenClaw, run these commands before doing anything else:

# Validate config and auto-fix common issues
openclaw doctor --fix

# Run a deep security scan
openclaw security audit --deep

The security audit should return zero critical issues. A couple of warnings is normal:

• gateway.trusted_proxies_missing — ok if localhost-only
• fs.credentials_dir.perms_readable — fix with chmod

File permissions

Lock down the config directory:

chmod 700 ~/.openclaw
chmod 600 ~/.openclaw/openclaw.json
chmod 700 ~/.openclaw/credentials

This prevents other users on the system from reading your config or API keys.

Gateway binding

Verify the gateway is binding to localhost and not exposed to the public internet:

netstat -an | grep 18789 | grep LISTEN
# You want to see: 127.0.0.1:18789
# You do NOT want: 0.0.0.0:18789

Fix it in config:

"gateway": {
  "bind": "loopback"
}

Hardening configuration

Logging with redaction

"logging": {
  "redactSensitive": "tools"
}

Tool policies

Restrict which tools agents can use:

"tools": {
  "profile": "minimal",
  "deny": ["exec"],
  "allow": ["web_search", "web_fetch", "read"]
}

Sandbox mode (optional)

If you want containerized execution (requires Docker):

"agents": {
  "defaults": {
    "sandbox": {
      "enabled": true,
      "image": "openclaw-sandbox"
    }
  }
}

Useful if running on a shared VPS and want agent work isolated.

Git-track your config

Git-track the OpenClaw config directory for rollback capability:

cd ~/.openclaw && git init
printf 'agents/*/sessions/\nagents/*/agent/*.jsonl\n*.log\n' > .gitignore
git add .gitignore openclaw.json
git commit -m "config: baseline"

# Before making changes
git commit -am "config: before model update"

# Make your changes
vim openclaw.json

# Test and commit
openclaw doctor --fix
git commit -am "config: switched to Gemini 3 Flash"

When something breaks at midnight, git diff and git checkout are faster than trying to remember what you changed.

Validation workflow

After any config change:

System monitoring

# Check OpenClaw status
openclaw status

# Check for errors in logs
tail -100 ~/.openclaw/gateway.log | grep -i error

# Check system resources
htop

Consider adding a cron job or heartbeat check for failed jobs and error logs — see the Heartbeat & Task Tracking chapter.

Backup strategy

Backup script

Create ~/bin/backup-openclaw.sh:

#!/bin/bash
# OpenClaw backup script
BACKUP_DIR=~/backups
DATE=$(date +%Y-%m-%d)

# Create backup directory if it doesn't exist
mkdir -p $BACKUP_DIR

# Create backup
tar czf $BACKUP_DIR/openclaw-$DATE.tar.gz \
  ~/.openclaw/openclaw.json \
  ~/.openclaw/credentials \
  ~/.openclaw/workspace

# Keep only last 7 days of backups
find $BACKUP_DIR -name "openclaw-*.tar.gz" -mtime +7 -delete

# Log completion
echo "$(date): Backup completed - openclaw-$DATE.tar.gz" >> $BACKUP_DIR/backup.log

chmod +x ~/bin/backup-openclaw.sh
crontab -e

# Daily at 3 AM
0 3 * * * ~/bin/backup-openclaw.sh

# Every 6 hours
0 */6 * * * ~/bin/backup-openclaw.sh

# Weekly on Sunday at 2 AM
0 2 * * 0 ~/bin/backup-openclaw.sh

Service management

The OpenClaw wizard automatically sets up a systemd user service on Linux (or launchd on macOS) during installation. No manual configuration needed.

# Check service status
openclaw status
# or
systemctl --user status openclaw

# Manage the service
openclaw gateway start
openclaw gateway stop
openclaw gateway restart

Get things stable before going 24/7

Don't run OpenClaw 24/7 until you've ironed out the basics. Get your config right, get your agents stable, get your costs predictable. Then let it run.

I've seen people jump to 24/7 operation on day one and end up with a bot that burns $50 in API calls overnight because a retry loop went sideways.

Troubleshooting

sudo tailscale status
sudo tailscale status | grep ssh
sudo tailscale up --ssh=true

tailscale ip -4    # IPv4
tailscale ip -6    # IPv6
hostname           # Machine name on Tailscale network

openclaw doctor --fix
openclaw gateway stop
openclaw gateway start

lsof -i :18789
# Kill the process or change the port in config

Pre-production security checklist

Cost optimization

Running on a VPS doesn't mean expensive:

• Hetzner CX23: ~$5/month
• Cheap model for heartbeat (GPT-5 nano)
• Cross-provider fallbacks to avoid quota burnout
• Concurrency limits to prevent runaway costs

See the Agent Prompts chapter for model configuration strategies.